Search Results

Search found 2899 results on 116 pages for 'rate limiting'.

Page 109/116 | < Previous Page | 105 106 107 108 109 110 111 112 113 114 115 116  | Next Page >

• CodePlex Daily Summary for Friday, August 15, 2014

  CodePlex Daily Summary for Friday, August 15, 2014Popular ReleasesGoogle .Net API: Drive.Sample: Google .NET Client API – Drive.SampleInstructions for the Google .NET Client API – Drive.Sample</h2> http://code.google.com/p/google-api-dotnet-client/source/browse/?repo=samples#hg%2FDrive.SampleBrowse Source, or main file http://code.google.com/p/google-api-dotnet-client/source/browse/Drive.Sample/Program.cs?repo=samplesProgram.cs <h3>1. Checkout Instructions</h3> <p><b>Prerequisites:</b> Install Visual Studio, and <a href="http://mercurial.selenic.com/">Mercurial</a>.</p> ...FineUI - jQuery / ExtJS based ASP.NET Controls: FineUI v4.1.1: -??Form??????????????（???-5929）。 -?TemplateField??ExpandOnDoubleClick、ExpandOnEnter、ExpandToSelectRow????（LZOM-5932）。 -BodyPadding???????，??“5”“5 10”，???????????“5px”“5px 10px”。 -??TriggerBox?EnableEdit=false????，??????????????（Jango_Jing-5450）。 -???????????DataKeyNames???????????（yygy-6002）。 -????????????????????????（Gnid-6018）。 -??PageManager???AutoSizePanelID????，??????????????????（yygy-6008）。 -?FState???????????????，????????????????（????-5925）。 -??????OnClientClick???return?????????（FineU...SEToolbox: SEToolbox 01.042.020 Release 1: Updated Mod support. On startup, only stock items will appear in the Components list. Upon selecting and loading a saved world, the mods for that world only will then be loaded, and only from the local drive. If a mod has not been downloaded in Space Engineers, it will not download it for you. If you are developing a Mod, hitting "Reload" will also reload the mods as well as the saved world. If SEToolbox is crashing when loading a saved world containing mods, it is most likely because one ...Gum UI Tool: Gum 0.6.09: Fixed bug which would not allow plugins to be loaded when the app was distributed. Added animation plugin7zbackup - PowerShell Script to Backup Files with 7zip: 7zBackup v. 1.9.8 Stable: Do you like this piece of software ? It took some time and effort to develop. Please consider helping me with a donation Feat : Lock file now holds process ID and RootDir. On subsequent launches script checks if previous process is still alive. In case it is not it will clean up orphaned junction root directory. Ensure no orphaned rootdirs are on disk and no lockfiles in %temp% directory before running this releaseDNN CMS Platform: 07.03.02: Major Highlights Fixed backwards compatibility issue with 3rd party control panels Fixed issue in the drag and drop functionality of the File Uploader in IE 11 and Safari Fixed issue where users were able to create pages with the same name Fixed issue that affected older versions of DNN that do not include the maxAllowedContentLength during upgrade Fixed issue that stopped some skins from being upgraded to newer versions Fixed issue that randomly showed an unexpected error during us...WordMat: WordMat for Mac: WordMat for Mac has a few limitations compared to the Windows version - Graph is not supported (Gnuplot, GeoGebra and Excel works) - Units are not supported yet (Coming up) The Mac version is yet as tested as the windows version.ConEmu - Windows console with tabs: ConEmu 140814 [Alpha]: ConEmu - developer build x86 and x64 versions. Written in C++, no additional packages required. Run "ConEmu.exe" or "ConEmu64.exe". Some useful information you may found: http://superuser.com/questions/tagged/conemu http://code.google.com/p/conemu-maximus5/wiki/ConEmuFAQ http://code.google.com/p/conemu-maximus5/wiki/TableOfContents If you want to use ConEmu in portable mode, just create empty "ConEmu.xml" file near to "ConEmu.exe" HP OneView PowerShell Library: HP OneView PowerShell Library 1.10.1193: Branch to HP OneView 1.10 Release. NOTE: This library version does not support older appliance versions. Fixed New-HPOVProfile to check for Firmware and BIOS management for supported platforms. Would erroneously error when neither -firmware or -bios were passed. Fixed Remove-HPOV* cmdlets which did not handle -force switch parameter correctly Fixed New-HPOVUplinkSet and New-HPOVNetwork Fixed Download-File where HTTP stream compression was not handled, resulting in incorrectly writt...NeoLua (Lua for .net dynamic language runtime): NeoLua-0.8.17: Fix: table.insert Fix: table auto convert Fix: Runtime-functions were defined as private it should be internal. Fix: min,max MichaelSenko release.MFCMAPI: August 2014 Release: Build: 15.0.0.1042 Full release notes at SGriffin's blog. If you just want to run the MFCMAPI or MrMAPI, get the executables. If you want to debug them, get the symbol files and the source. The 64 bit builds will only work on a machine with Outlook 2010/2013 64 bit installed. All other machines should use the 32 bit builds, regardless of the operating system. Facebook BadgeOooPlayer: 1.1: Added: Support for speex, TAK and OptimFrog files Added: An option to not to load cover art Added: Smaller package size Fixed: Unable to drag&drop audio files to playlist Updated: FLAC, WacPack and Opus playback libraries Updated: ID3v1 and ID3v2 tag librariesEWSEditor: EwsEditor 1.10 Release: • Export and import of items as a full fidelity steam works - without proxy classes! - I used raw EWS POSTs. • Turned off word wrap for EWS request field in EWS POST windows. • Several windows with scrolling texts boxes were limiting content to 32k - I removed this restriction. • Split server timezone info off to separate menu item from the timezone info windows so that the timezone info window could be used without logging into a mailbox. • Lots of updates to the TimeZone window. • UserAgen...Python Tools for Visual Studio: 2.1 RC: Release notes for PTVS 2.1 RC We’re pleased to announce the release candidate for Python Tools for Visual Studio 2.1. Python Tools for Visual Studio (PTVS) is an open-source plug-in for Visual Studio which supports programming with the Python language. PTVS supports a broad range of features including CPython/IronPython, editing, IntelliSense, interactive debugging, profiling, Microsoft Azure, IPython, and cross-platform debugging support. PTVS 2.1 RC is available for: Visual Studio Expre...Sense/Net ECM - Enterprise CMS: SenseNet 6.3.1 Community Edition: Sense/Net 6.3.1 Community EditionSense/Net 6.3.1 is an important step toward a more modular infrastructure, robustness and maintainability. With this release we finally introduce a packaging and a task management framework, and the Image Editor that will surely make the job of content editors more fun. Please review the changes and new features since Sense/Net 6.3 and give a feedback on our forum! Main new featuresSnAdmin (packaging framework) Task Management Image Editor OData REST A...Fluffy: Fluffy 0.3.35.4: Change log: Text editorSKGL - Serial Key Generating Library: SKGL Extension Methods 4 (1.0.5.1): This library contains methods for: Time change check (make sure the time has not been changed on the client computer) Key Validation (this will use http://serialkeymanager.com/ to validate keys against the database) Key Activation (this will, depending on the settings, activate a key with a specific machine code) Key Activation Trial (allows you to update a key if it is a trial key) Get Machine Code (calculates a machine code given any hash function) Get Eight Byte Hash (returns an...Touchmote: Touchmote 1.0 beta 13: Changes Less GPU usage Works together with other Xbox 360 controls Bug fixesModern UI for WPF: Modern UI 1.0.6: The ModernUI assembly including a demo app demonstrating the various features of Modern UI for WPF. BREAKING CHANGE LinkGroup.GroupName renamed to GroupKey NEW FEATURES Improved rendering on high DPI screens, including support for per-monitor DPI awareness available in Windows 8.1 (see also Per-monitor DPI awareness) New ModernProgressRing control with 8 builtin styles New LinkCommands.NavigateLink routed command New Visual Studio project templates 'Modern UI WPF App' and 'Modern UI W...ClosedXML - The easy way to OpenXML: ClosedXML 0.74.0: Multiple thread safe improvements including AdjustToContents XLHelper XLColor_Static IntergerExtensions.ToStringLookup Exception now thrown when saving a workbook with no sheets, instead of creating a corrupt workbook Fix for hyperlinks with non-ASCII Characters Added basic workbook protection Fix for error thrown, when a spreadsheet contained comments and images Fix to Trim function Fix Invalid operation Exception thrown when the formula functions MAX, MIN, and AVG referenc...New Projectsapple TV: Apple TV project homepageArma 3 Battle Eye Client: Arma3BEClientASP.NET MVC AngularJS w/ Google Maps API: ASP.NET MVC sample using Google Maps API w/ AngularJS.CC-Classwork: Classwork from CoderCampsCompanyPortal: CompanyPortalcore: Building an Internet of Things (IoT, also Cloud of Things or CoT) core, drawing inspirations from the pre-existing Linus Torvalds linux kernel made from GNU/nixCRM Early Bound Class Simplifier: Simplifies the creation of a Dynamics CRM Early Bound Class. Dirección Desconcentrada de Cultura: Este proyecto web se ha elaborado para la dirección desconcentrada de cultura de cajamarca a cargo de los practicantes de UPNC Sitemas computacionales.Energy Trail Site: NGO Site for designing and collaboration work.Hybrid Platform - Build anything: A Platform that built by loosely coupled architecture. You can build applications for Web, Desktop, Mobile, WCF Services - ASP.NET MVC on this concrete platformipad air: a web tool to sim display same as ipad airipad apps: A serices to support Ipad HD devise to request CURD for codeplex.comiphone 6: iphone6iphone air: Opend API lists for IPhone 6(iphone air)iphone apps: Bus API for iphoneiwatch: A priview version for iwtach API Named Colors in Silverlight: This project is a Silverlight dll to add the missing named colors from System.Windows.Media.Color. Once added as a reference, it makes using named colors easy!OOP_2113110295: Name: Nguyen Trung Thao ID 2113110295 Truong Cao Dang Cong Thuong Mon: OOPPagepark: PageparkProjektRepository: Eine virtuelle Forschungsumgebung (VFU) um Forschungsdaten und Artefakte zu sammeln, gemeinsam zu nutzen, erschließen und mit Metadaten anreichern zu könnenRamonaSniffer: This will be the repository to host the zigbee snifferseawol: A Blog system base on node.jsSonar settings for TFS Build: Sample of configurations for Sonar to work with TFS for copy/pasteSon's Homework and learning to code: Just a collection of coding projects to learn from.SunBurn Terrain Editor: A fully functional standalone WYSWYG terrain (height map and color map) editor. Built upon the SunBurn Platform Framework allowing scope for Linux and Mac ports????.????????: 1) ??????? ???????? ?? 2) C# ?????????? (??????) ??? ???????? ?????? ???? (? ??????? *.dbf) ? ????? ???? 3) WinForms-?????????? ??? ???????????? ?????? ????

  Read the article

• 10 tape technology features that make you go hmm.

  - by Karoly Vegh

  A week ago an Oracle/StorageTek Tape Specialist, Christian Vanden Balck, visited Vienna, and agreed to visit customers to do techtalks and update them about the technology boom going around tape. I had the privilege to attend some of his sessions and noted the information and features that took the customers by surprise and made them think. Allow me to share the top 10: I. StorageTek as a brand: StorageTek is one of he strongest names in the Tape field. The brand itself was valued so much by customers that even after Sun Microsystems acquiring StorageTek and the Oracle acquiring Sun the brand lives on with all the Oracle tapelibraries are officially branded StorageTek.See http://www.oracle.com/us/products/servers-storage/storage/tape-storage/overview/index.html II. Disk information density limitations: Disk technology struggles with information density. You haven't seen the disk sizes exploding lately, have you? That's partly because there are physical limits on a disk platter. The size is given, the number of platters is limited, they just can't grow, and are running out of physical area to write to. Now, in a T10000C tape cartridge we have over 1000m long tape. There you go, you have got your physical space and don't need to stuff all that data crammed together. You can write in a reliable pattern, and have space to grow too. III. Oracle has a market share of 62% worldwide in recording head manufacturing. That's right. If you are running LTO drives, with a good chance you rely on StorageTek production. That's two out of three LTO recording heads produced worldwide.  IV. You can store 1 Exabyte data in a single tape library. Yes, an Exabyte. That is 1000 Petabytes. Or, a million Terabytes. A thousand million GigaBytes. You can store that in a stacked StorageTek SL8500 tapelibrary. In one SL8500 you can put 10.000 T10000C cartridges, that store 10TB data (compressed). You can stack 10 of these SL8500s together. Boom. 1000.000 TB.(n.b.: stacking means interconnecting the libraries. Yes, cartridges are moved between the stacked libraries automatically.)  V. EMC: 'Tape doesn't suck after all. We moved on.': Do you remember the infamous 'Tape sucks, move on' Datadomain slogan? Of course they had to put it that way, having only had disk products. But here's a fun fact: on the EMCWorld 2012 there was a major presence of a Tape-tech company - EMC, in a sudden burst of sanity is embracing tape again. VI. The miraculous T10000C: Oracle StorageTek has developed an enterprise-grade tapedrive and cartridge, the T10000C. With awesome numbers: The Cartridge: Native 5TB capacity, 10TB with compression Over a kilometer long tape within the cartridge. And it's locked when unmounted, no rattling of your data.  Replaced the metalparticles datalayer with BaFe (bariumferrite) - metalparticles lose around 7% of magnetism within 30 days. BaFe does not. Yes we employ solid-state physicists doing R&D on demagnetisation in our labs. Can be partitioned, storage tiering within the cartridge!  The Drive: 2GB Cache Encryption implemented in HW - no performance hit 252 MB/s native sustained data rate, beats disk technology by far. Not to mention peak throughput.  Leading the tape while never touching the data side of it, protecting your data physically too Data integritiy checking (CRC recalculation) on tape within the drive without having to read it back to the server reordering data from tape-order, delivering it back in application-order  writing 32 tracks at once, reading them back for CRC check at once VII. You only use 20% of your data on a regular basis. The rest 80% is just lying around for years. On continuously spinning disks. Doubly consuming energy (power+cooling), blocking diskstorage capacity. There is a solution called SAM (Storage Archive Manager) that provides you a filesystem unifying disk and tape, moving data on-demand and for clients transparently between the different storage tiers. You can share these filesystems with NFS or CIFS for clients, and enjoy the low TCO of tape. Tapes don't spin. They sit quietly in their slots, storing 10TB data, using no energy, producing no heat, automounted when a client accesses their data.See: http://www.oracle.com/us/products/servers-storage/storage/storage-software/storage-archive-manager/overview/index.html VIII. HW supported for three decades: Did you know that the original PowderHorn library was released in '87 and has been only discontinued in 2010? That is over two decades of supported operation. Tape libraries are - just like the data carrying on tapecartridges - built for longevity. Oh, and the T10000C cartridge has 30-year archival life for long-term retention.  IX. Tape is easy to manage: Have you heard of Tape Storage Analytics? It is a central graphical tool to summarize, monitor, analyze dataflow, health and performance of drives and libraries, see: http://www.oracle.com/us/products/servers-storage/storage/tape-storage/tape-analytics/overview/index.html X. The next generation: The T10000B drives were able to reuse the T10000A cartridges and write on them even more data. On the same cartridges. We call this investment protection, and this is very important for Oracle for the future too. We usually support two generations of cartridges together. The current drive is a T10000C. (...I know I promised to enlist 10, but I got still two more I really want to mention. Allow me to work around the problem: ) X++. The TallBots, the robots moving around the cartridges in the StorageTek library from tapeslots to the drives are cableless. Cables, belts, chains running to moving parts in a library cause maintenance downtimes. So StorageTek eliminated them. The TallBots get power, commands, even firmwareupgrades through the rails they are running on. Also, the TallBots don't just hook'n'pull the tapes out of their slots, they actually grip'n'lift them out. No friction, no scratches, no zillion little plastic particles floating around in the library, in the drives, on your data. (X++)++: Tape beats SSDs and Disks. In terms of throughput (252 MB/s), in terms of TCO: disks cause around 290x more power and cooling, in terms of capacity: 10TB on a single media and soon more.  So... do you need to store large amounts of data? Are you legally bound to archive it for dozens of years? Would you benefit from automatic storage tiering? Have you got large mediachunks to be streamed at times? Have you got power and cooling issues in the growing datacenters? Do you find EMC's 180° turn of tape attitude interesting, but appreciate it at the same time? With all that, you aren't alone. The most data on this planet is stored on tape. Tape is coming. Big time.

  Read the article

• first install for windows eight.....da beta

  - by raysmithequip

  The W8 preview is now installed and I am enjoying it.  I remember the learning curve of my first unix machine back in the eighties, this ain't that.It is normal for me to do the first os install with a keyboard and low end monitor...you never know what you'll encounter out in the field.  The OS took like a fish to water.  I used a low end INTEL motherboard dp55w I gathered on the cheap, an 1157 i5 from the used bin a pair of 6 gig ddr3 sticks, a rosewell 550 watt power supply a cheap used twenty buck sub 200g wd sata drive, a half working dvd burner and an asus fanless nvidia vid card, not a great one but Sub 50.00 on newey eggey...I did have to hunt the ms forums for a key and of course to activate the thing, if dos would of needed this outmoded ritual, we would still be on cpm and osborne would be a household name, of course little do people know that this ritual was common as far back as the seventies on att unix installs....not, but it was possible, I used to joke about when I ran a bbs, what hell would of been wrought had dos 3.2 machines been required to dial into my bbs to send fido mail to ms and wait for an acknowledgement.  All in all the thing was pushing a seven on the ms richter scale, not including the vid card, sadly it came in at just a tad over three....I wanted to evaluate it for a possible replacement on critical machines that in the past went down due to a vid card fan failure....you have no idea what a customer thinks when you show them a failed vid card fan..."you mean that little plastic piece of junk caused all this!!??!!!"...yea man.  Some production machines don't need any sort of vid, I will at least keep it on the maybe list for those, MTBF is a very important factor, some big box stores should put percentage of failure rate within 24 month estimates on the outside of the carton for sure.  And a warning that the power supplies are already at their limit.  Let's face it, today even 550w can be iffy.A few neat eye candy improvements over the earlier windows is nice, the metro screen is nice, anyone who has used a newer phone recently will intuitively drag their fingers across the screen....lot of good that was with no mouse or touch screen though.  Lucky me, I have been using windows since day one, I still have a copy of win 2.0 (and every other version) for no good reason.  Still the old ix collection of disks is much larger, recompiling any kernal is another silly ritual, same machine, different day, same recompile...argh. Rh is my all time fav, mandrake was always missing something, like it rewrote the init file or something, novell is ok as long as you stay on the beaten path and of course ubuntu normally recompiles with the same errors consistantly....makes life easy that way....no errors on windows eight, just a screen that did not match the installed hardware, natuarally I alt tabbed right out of it, then hit the flag key to find the start menu....no start button. I miss the start button already. Keyboard cowboy funnin and I was browsing the harddrive, nothing stunning there, I like that, means I can find stuff. Only I can't find what I want, the start button....the start menu is that first screen for touch tablets. No biggie for useruser, that is where they will want to be, I can see that. Admins won't want to be there, it is easy enough to get the control panel a bazzilion other ways though, just not the start button. (see a pattern here?). Personally, from the keyboard I find it fun to hit the carets along the location bar at the top of the explorer screen with tabs and arrows and choose SHOW ALL CONTROL PANEL ITEMS, or thereabouts. Bottom line, I love seven and I'll love eight even more!...very happy I did not have to follow the normal rule of thumb (a customer watching me build a system and asking questions said "oh I get it, so every piece you put in there is basically a hundred bucks, right?)...ok, sure, pretty much, more or less, well, ya dude.  It will be WAY past october till I get a real touch screen but I did pick up a pair of cheap tatungs so I can try the NEW main start screen, I parse a lot of folders and have a vision of how a pair of touch screens will be easier than landing a rover on mars.  Ok.  fine, they are way smallish, and I don't expect multitouch to work but we are talking a few percent of a new 21 inch viewsonic touch screen.  Will this OS be a game changer?  I don't know.  Bottom line with all the pads and droids in the world, it is more of a catch up move at first glance.  Not something ms is used to.  An app store?  I can see ms's motivation, the others have it.  I gather there will not be gadgets there, go ahead and see what ms did  to the once populated gadget page...go ahead, google gadgets and take a gander, used to hundreds of gadgets, they are already gone.  They replaced gadgets?  sort of, I'll drop that, it's a bit of a sore point for me.  More of interest was what happened when I downloaded stuff off codeplex and some other normal programs that I like, like orbitron, top o' my list!!...cardware it is...anyways, click on the exe, get a screen, normal for windows, this one indicated that I was not running a normal windows program and had a button for  exit the install, naw, I hit details, a hidden run program anyways came into view....great, my path to the normal windows has detected a program tha.....yea ok, acl is on, fine, moving along I got orbitron installed in record time and was tracking the iss on the newest Microsoft OS, beta of course, felt like the first time I setup bsd all those year ago...FUN!!...I suppose I gotta start to think about budgeting for the real os when it comes out in october, by then I should have a rasberry pi and be done with fedora remixed.  Of course that sounds like fun too!!  I would use this OS on a tablet or phone.  I don't like the idea of being hearded to an app store, don't like that on anything, we are americans and want real choices not marketed hype, lest you are younger with opm (other peoples money).   This os would be neat on a zune, but I suspect the zune is a gonner, I am rooting for microsoft, after all their default password is not admin anymore, nor alpine,  it's blank. Others force a password, my first fawn password was so long I could not even log into it with the password in front of me, who the heck uses %$# anyways, and if I was writing a brute force attack what the heck kinda impasse is that anyways at .00001 microseconds of a code execution cycle (just a non qualified number, not a real clock speed)....AI is where it will be before too long, MS is on that path, perhaps soon someone will sit down and write an app for the kinect that watches your eyes while you scan the new main start screen, clicking on the big E icon when you blink.....boy is that going to be fun!!!! sure. Blink,dammit,blink,dammit...... OPM no doubt.I like windows eight, we are moving forwards, better keep a close eye on ubuntu.  The real clinch comes when open source becomes paid source......don't blink, I already see plenty of very expensive 'ix apps, some even in app stores already.  more to come.......

  Read the article

• NUMA-aware placement of communication variables

  - by Dave

  For classic NUMA-aware programming I'm typically most concerned about simple cold, capacity and compulsory misses and whether we can satisfy the miss by locally connected memory or whether we have to pull the line from its home node over the coherent interconnect -- we'd like to minimize channel contention and conserve interconnect bandwidth. That is, for this style of programming we're quite aware of where memory is homed relative to the threads that will be accessing it. Ideally, a page is collocated on the node with the thread that's expected to most frequently access the page, as simple misses on the page can be satisfied without resorting to transferring the line over the interconnect. The default "first touch" NUMA page placement policy tends to work reasonable well in this regard. When a virtual page is first accessed, the operating system will attempt to provision and map that virtual page to a physical page allocated from the node where the accessing thread is running. It's worth noting that the node-level memory interleaving granularity is usually a multiple of the page size, so we can say that a given page P resides on some node N. That is, the memory underlying a page resides on just one node. But when thinking about accesses to heavily-written communication variables we normally consider what caches the lines underlying such variables might be resident in, and in what states. We want to minimize coherence misses and cache probe activity and interconnect traffic in general. I don't usually give much thought to the location of the home NUMA node underlying such highly shared variables. On a SPARC T5440, for instance, which consists of 4 T2+ processors connected by a central coherence hub, the home node and placement of heavily accessed communication variables has very little impact on performance. The variables are frequently accessed so likely in M-state in some cache, and the location of the home node is of little consequence because a requester can use cache-to-cache transfers to get the line. Or at least that's what I thought. Recently, though, I was exploring a simple shared memory point-to-point communication model where a client writes a request into a request mailbox and then busy-waits on a response variable. It's a simple example of delegation based on message passing. The server polls the request mailbox, and having fetched a new request value, performs some operation and then writes a reply value into the response variable. As noted above, on a T5440 performance is insensitive to the placement of the communication variables -- the request and response mailbox words. But on a Sun/Oracle X4800 I noticed that was not the case and that NUMA placement of the communication variables was actually quite important. For background an X4800 system consists of 8 Intel X7560 Xeons . Each package (socket) has 8 cores with 2 contexts per core, so the system is 8x8x2. Each package is also a NUMA node and has locally attached memory. Every package has 3 point-to-point QPI links for cache coherence, and the system is configured with a twisted ladder "mobius" topology. The cache coherence fabric is glueless -- there's not central arbiter or coherence hub. The maximum distance between any two nodes is just 2 hops over the QPI links. For any given node, 3 other nodes are 1 hop distant and the remaining 4 nodes are 2 hops distant. Using a single request (client) thread and a single response (server) thread, a benchmark harness explored all permutations of NUMA placement for the two threads and the two communication variables, measuring the average round-trip-time and throughput rate between the client and server. In this benchmark the server simply acts as a simple transponder, writing the request value plus 1 back into the reply field, so there's no particular computation phase and we're only measuring communication overheads. In addition to varying the placement of communication variables over pairs of nodes, we also explored variations where both variables were placed on one page (and thus on one node) -- either on the same cache line or different cache lines -- while varying the node where the variables reside along with the placement of the threads. The key observation was that if the client and server threads were on different nodes, then the best placement of variables was to have the request variable (written by the client and read by the server) reside on the same node as the client thread, and to place the response variable (written by the server and read by the client) on the same node as the server. That is, if you have a variable that's to be written by one thread and read by another, it should be homed with the writer thread. For our simple client-server model that means using split request and response communication variables with unidirectional message flow on a given page. This can yield up to twice the throughput of less favorable placement strategies. Our X4800 uses the QPI 1.0 protocol with source-based snooping. Briefly, when node A needs to probe a cache line it fires off snoop requests to all the nodes in the system. Those recipients then forward their response not to the original requester, but to the home node H of the cache line. H waits for and collects the responses, adjudicates and resolves conflicts and ensures memory-model ordering, and then sends a definitive reply back to the original requester A. If some node B needed to transfer the line to A, it will do so by cache-to-cache transfer and let H know about the disposition of the cache line. A needs to wait for the authoritative response from H. So if a thread on node A wants to write a value to be read by a thread on node B, the latency is dependent on the distances between A, B, and H. We observe the best performance when the written-to variable is co-homed with the writer A. That is, we want H and A to be the same node, as the writer doesn't need the home to respond over the QPI link, as the writer and the home reside on the very same node. With architecturally informed placement of communication variables we eliminate at least one QPI hop from the critical path. Newer Intel processors use the QPI 1.1 coherence protocol with home-based snooping. As noted above, under source-snooping a requester broadcasts snoop requests to all nodes. Those nodes send their response to the home node of the location, which provides memory ordering, reconciles conflicts, etc., and then posts a definitive reply to the requester. In home-based snooping the snoop probe goes directly to the home node and are not broadcast. The home node can consult snoop filters -- if present -- and send out requests to retrieve the line if necessary. The 3rd party owner of the line, if any, can respond either to the home or the original requester (or even to both) according to the protocol policies. There are myriad variations that have been implemented, and unfortunately vendor terminology doesn't always agree between vendors or with the academic taxonomy papers. The key is that home-snooping enables the use of a snoop filter to reduce interconnect traffic. And while home-snooping might have a longer critical path (latency) than source-based snooping, it also may require fewer messages and less overall bandwidth. It'll be interesting to reprise these experiments on a platform with home-based snooping. While collecting data I also noticed that there are placement concerns even in the seemingly trivial case when both threads and both variables reside on a single node. Internally, the cores on each X7560 package are connected by an internal ring. (Actually there are multiple contra-rotating rings). And the last-level on-chip cache (LLC) is partitioned in banks or slices, which with each slice being associated with a core on the ring topology. A hardware hash function associates each physical address with a specific home bank. Thus we face distance and topology concerns even for intra-package communications, although the latencies are not nearly the magnitude we see inter-package. I've not seen such communication distance artifacts on the T2+, where the cache banks are connected to the cores via a high-speed crossbar instead of a ring -- communication latencies seem more regular.

  Read the article

• TGIF: Engagement Wrap-up

  - by Michael Snow

  We've had a very busy week here at Oracle and as we build up to Oracle OpenWorld starting in less than 10 days - it doesn't look like things will be slowing down. Engagement is definitely in the air this week. Our friend, John Mancini published a great article entitled: "The World of Engagement" on his Digital Landfill blog yesterday and we hosted a great webcast with R "Ray" Wang from Constellation Research yesterday on the "9 C's of Engagement". 12.00 Normal 0 false false false EN-US X-NONE X-NONE MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-family:"Calibri","sans-serif"; mso-bidi-font-family:"Times New Roman";} I wanted to wrap-up the week with some key takeaways from our webcast yesterday with Ray Wang. If you missed the webcast yesterday, fear not - it is now available  On-Demand. We'll leave you this week with lots of questions about how to navigate these churning waters of engagement. Stay tuned to the Oracle WebCenter Social Business Thought Leaders Webcast Series as we fuel this dialogue. 12.00 Normal 0 false false false EN-US X-NONE X-NONE MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Calibri","sans-serif"; mso-bidi-font-family:"Times New Roman";} Company Culture Does company support a culture of putting customer satisfaction ahead of profits? Does culture promote creativity and cross functional employee collaboration? Does culture accept different views of multi-generational workforce? Does culture promote employee training and skills development Does culture support upward mobility and long term retention? Does culture support work-life balance? Does the culture provide rewards for employee for outstanding customer support? Channels What are the current primary channels for customer communications? What do you think will be the primary channels in two years? Is company developing support model for emerging channels? Do all channels consistently deliver the same level of customer support? Do you know the cost per transaction across all channels? Do you engage customers proactively across multiple channels? Do all channels have access to the same customer information? Community Does company extend customer support into virtual communities of interest? Does company facilitate educating users through its virtual communities? Does company mine its customer’s experience into useful data? Does company increase the value for customers through using data to deliver new products and services? Does company support two way interactions with its customers through communities of interest? Does company actively support social CRM, online communities and social media markets? Credibility Does company market its trustworthiness through external certificates such as business licenses, BBB certificates or other validations? Does company promote trust through customer testimonials and case studies on ethical business practices? Does company promote truthful market campaigns Does company make it easy for customers to complain? Does company build its reputation for standing behind its products with guarantees for satisfaction? Does company protect its customer data with high security measures> Content What sources do you use to create customer content? Does company mine social media and blogs for customer content? How does your company sort, store and retain its customer content? How frequently does content get updated? What external sources do you use for customer content? How many responses are typically received from a knowledge management system inquiry? Does your company use customer content to design and develop new product and services? Context Does your company market to customers in clusters or individually? Does your company customize its messages and personalize them to specific needs of each individual customer? Does your company store customer data based on their past behaviors, purchases, sentiment analysis and current activities? Does your company manage customer context according to channels used? For example identify personal use channels versus business channels? What is your frequency of collecting customer activities across various touch points? How is your customer data stored and analyzed? Is contextual data used for future customer outreach? Cadence Which channels does your company measure-web site visits, phone calls, IVR, store visits, face to face, social media? Does company make effective use of cross channel marketing to promote more frequent customer engagement? Does your company rate the patterns relevant for your product or service and monitor usage against this pattern? Does your company measure the frequency of both online and offline channels? Does your company apply metrics to the frequency of customer engagements with product or services revenues? Does your company consolidate data for customer engagement across various channels for a complete view of its customer? Catalyst Does company offer coupon discounts? Does company have a customer loyalty program or a VIP membership program? Does company mine customer data to target specific groups of buyers? Do internal employees serve as ambassadors for customer programs? Does company drive loyalty through social media loyalty programs? Does company build rewards based on using loyalty data? Does company offer an employee incentive program to drive customer loyalty?

  Read the article

• Augmenting your Social Efforts via Data as a Service (DaaS)

  - by Mike Stiles

  The following is the 3rd in a series of posts on the value of leveraging social data across your enterprise by Oracle VP Product Development Don Springer and Oracle Cloud Data and Insight Service Sr. Director Product Management Niraj Deo. In this post, we will discuss the approach and value of integrating additional “public” data via a cloud-based Data-as-as-Service platform (or DaaS) to augment your Socially Enabled Big Data Analytics and CX Management. Let’s assume you have a functional Social-CRM platform in place. You are now successfully and continuously listening and learning from your customers and key constituents in Social Media, you are identifying relevant posts and following up with direct engagement where warranted (both 1:1, 1:community, 1:all), and you are starting to integrate signals for communication into your appropriate Customer Experience (CX) Management systems as well as insights for analysis in your business intelligence application. What is the next step? Augmenting Social Data with other Public Data for More Advanced Analytics When we say advanced analytics, we are talking about understanding causality and correlation from a wide variety, volume and velocity of data to Key Performance Indicators (KPI) to achieve and optimize business value. And in some cases, to predict future performance to make appropriate course corrections and change the outcome to your advantage while you can. The data to acquire, process and analyze this is very nuanced: It can vary across structured, semi-structured, and unstructured data It can span across content, profile, and communities of profiles data It is increasingly public, curated and user generated The key is not just getting the data, but making it value-added data and using it to help discover the insights to connect to and improve your KPIs. As we spend time working with our larger customers on advanced analytics, we have seen a need arise for more business applications to have the ability to ingest and use “quality” curated, social, transactional reference data and corresponding insights. The challenge for the enterprise has been getting this data inline into an easily accessible system and providing the contextual integration of the underlying data enriched with insights to be exported into the enterprise’s business applications. The following diagram shows the requirements for this next generation data and insights service or (DaaS): Some quick points on these requirements: Public Data, which in this context is about Common Business Entities, such as - Customers, Suppliers, Partners, Competitors (all are organizations) Contacts, Consumers, Employees (all are people) Products, Brands This data can be broadly categorized incrementally as - Base Utility data (address, industry classification) Public Master Reference data (trade style, hierarchy) Social/Web data (News, Feeds, Graph) Transactional Data generated by enterprise process, workflows etc. This Data has traits of high-volume, variety, velocity etc., and the technology needed to efficiently integrate this data for your needs includes - Change management of Public Reference Data across all categories Applied Big Data to extract statics as well as real-time insights Knowledge Diagnostics and Data Mining As you consider how to deploy this solution, many of our customers will be using an online “cloud” service that provides quality data and insights uniformly to all their necessary applications. In addition, they are requesting a service that is: Agile and Easy to Use: Applications integrated with the service can obtain data on-demand, quickly and simply Cost-effective: Pre-integrated into applications so customers don’t have to Has High Data Quality: Single point access to reference data for data quality and linkages to transactional, curated and social data Supports Data Governance: Becomes more manageable and cost-effective since control of data privacy and compliance can be enforced in a centralized place Data-as-a-Service (DaaS) Just as the cloud has transformed and now offers a better path for how an enterprise manages its IT from their infrastructure, platform, and software (IaaS, PaaS, and SaaS), the next step is data (DaaS). Over the last 3 years, we have seen the market begin to offer a cloud-based data service and gain initial traction. On one side of the DaaS continuum, we see an “appliance” type of service that provides a single, reliable source of accurate business data plus social information about accounts, leads, contacts, etc. On the other side of the continuum we see more of an online market “exchange” approach where ISVs and Data Publishers can publish and sell premium datasets within the exchange, with the exchange providing a rich set of web interfaces to improve the ease of data integration. Why the difference? It depends on the provider’s philosophy on how fast the rate of commoditization of certain data types will occur. How do you decide the best approach? Our perspective, as shown in the diagram below, is that the enterprise should develop an elastic schema to support multi-domain applicability. This allows the enterprise to take the most flexible approach to harness the speed and breadth of public data to achieve value. The key tenet of the proposed approach is that an enterprise carefully federates common utility, master reference data end points, mobility considerations and content processing, so that they are pervasively available. One way you may already be familiar with this approach is in how you do Address Verification treatments for accounts, contacts etc. If you design and revise this service in such a way that it is also easily available to social analytic needs, you could extend this to launch geo-location based social use cases (marketing, sales etc.). Our fundamental belief is that value-added data achieved through enrichment with specialized algorithms, as well as applying business “know-how” to weight-factor KPIs based on innovative combinations across an ever-increasing variety, volume and velocity of data, will be where real value is achieved. Essentially, Data-as-a-Service becomes a single entry point for the ever-increasing richness and volume of public data, with enrichment and combined capabilities to extract and integrate the right data from the right sources with the right factoring at the right time for faster decision-making and action within your core business applications. As more data becomes available (and in many cases commoditized), this value-added data processing approach will provide you with ongoing competitive advantage. Let’s look at a quick example of creating a master reference relationship that could be used as an input for a variety of your already existing business applications. In phase 1, a simple master relationship is achieved between a company (e.g. General Motors) and a variety of car brands’ social insights. The reference data allows for easy sort, export and integration into a set of CRM use cases for analytics, sales and marketing CRM. In phase 2, as you create more data relationships (e.g. competitors, contacts, other brands) to have broader and deeper references (social profiles, social meta-data) for more use cases across CRM, HCM, SRM, etc. This is just the tip of the iceberg, as the amount of master reference relationships is constrained only by your imagination and the availability of quality curated data you have to work with. DaaS is just now emerging onto the marketplace as the next step in cloud transformation. For some of you, this may be the first you have heard about it. Let us know if you have questions, or perspectives. In the meantime, we will continue to share insights as we can.Photo: Erik Araujo, stock.xchng

  Read the article

• The True Cost of a Solution

  - by D'Arcy Lussier

  I had a Twitter chat recently with someone suggesting Oracle and SQL Server were losing out to OSS (Open Source Software) in the enterprise due to their issues with scaling or being too generic (one size fits all). I challenged that a bit, as my experience with enterprise sized clients has been different – adverse to OSS but receptive to an established vendor. The response I got was: Found it easier to influence change by showing how X can’t solve our problems or X is extremely costly to scale. Money talks. I think this is definitely the right approach for anyone pitching an alternate or alien technology as part of a solution: identify the issue, identify the solution, then present pros and cons including a cost/benefit analysis. What can happen though is we get tunnel vision and don’t present a full view of the costs associated with a solution. An “Acura”te Example (I’m so clever…) This is my dream vehicle, a Crystal Black Pearl coloured Acura MDX with the SH-AWD package! We’re a family of 4 (5 if my daughters ever get their wish of adding a dog), and I’ve always wanted a luxury type of vehicle, so this is a perfect replacement in a few years when our Rav 4 has hit the 8 – 10 year mark. MSRP – $62,890 But as we all know, that’s not *really* the cost of the vehicle. There’s taxes and fees added on, there’s the extended warranty if I choose to purchase it, there’s the finance rate that needs to be factored in… MSRP –   $62,890 Taxes –      $7,546 Warranty - $2,500 SubTotal – $72,936 Finance Charge – $ 1094.04 Grand Total – $74,030 Well! Glad we did that exercise – we discovered an extra $11k added on to the MSRP! Well now we have our true price…or do we? Lifetime of the Vehicle I’m expecting to have this vehicle for 7 – 10 years. While the hard cost of the vehicle is known and dealt with, the costs to run and maintain the vehicle are on top of this. I did some research, and here’s what I’ve found: Fuel and Mileage Gas prices are high as it is for regular fuel, but getting into an MDX will require that I *only* purchase premium fuel, which comes at a premium price. I need to expect my bill at the pump to be higher. Comparing the MDX to my 2007 Rav4 also shows I’ll be gassing up more often. The Rav4 has a city MPG of 21, while the MDX plummets to 16! The MDX does have a bigger fuel tank though, so all in all the number of times I hit the pumps might even out. Still, I estimate I’ll be spending approximately $8000 – $10000 more on gas over a 10 year period than my current Rav4. Service Options Limited Although I have options with my Toyota here in Winnipeg (we have 4 Toyota dealerships), I do go to my original dealer for any service work. Still, I like the fact that I have options. However, there’s only one Acura dealership in all of Winnipeg! So if, for whatever reason, I’m not satisfied with the level of service I’m stuck. Non Warranty Service Work Also let’s not forget that there’s a bulk of work required every year that is *not* covered under warranty – oil changes, tire rotations, brake pads, etc. I expect I’ll need to get new tires at the 5 years mark as well, which can easily be $1200 – $1500 (I just paid $1000 for new tires for the Rav4 and we’re at the 5 year mark). Now these aren’t going to be *new* costs that I’m not used to from our existing vehicles, but they should still be factored in. I’d budget $500/year, or $5000 over the 10 years I’ll own the vehicle. Final Assessment So let’s re-assess the true cost of my dream MDX: MSRP                    $62,890 Taxes                       $7,546 Warranty                 $2,500 Finance Charge         $1094 Gas                        $10,000 Service Work            $5000 Grand Total           $89,030 So now I have a better idea of 10 year cost overall, and I’ve identified some concerns with local service availability. And there’s now much more to consider over the original $62,890 price tag. Tying This Back to Technology Solutions The process that we just went through is no different than what organizations do when considering implementing a new system, technology, or technology based solution, within their environments. It’s easy to tout the short term cost savings of particular product/platform/technology in a vacuum. But its when you consider the wider impact that the true cost comes into play. Let’s create a scenario: A company is not happy with its current data reporting suite. An employee suggests moving to an open source solution. The selling points are: - Because its open source its free - The organization would have access to the source code so they could alter it however they wished - It provided features not available with the current reporting suite At first this sounds great to the management and executive, but then they start asking some questions and uncover more information: - The OSS product is built on a technology not used anywhere within the organization - There are no vendors offering product support for the OSS product - The OSS product requires a specific server platform to operate on, one that’s not standard in the organization All of a sudden, the true cost of implementing this solution is starting to become clearer. The company might save money on licensing costs, but their training costs would increase significantly – developers would need to learn how to develop in the technology the OSS solution was built on, IT staff must learn how to set up and maintain a new server platform within their existing infrastructure, and if a problem was found there was no vendor to contact for support. The true cost of implementing a “free” OSS solution is actually spinning up a project to implement it within the organization – no small cost. And that’s just the short-term cost. Now the organization must ensure they maintain trained staff who can make changes to the OSS reporting solution and IT staff that will stay knowledgeable in the new server platform. If those skills are very niche, then higher labour costs could be incurred if those people are hard to find or if trained employees use that knowledge as leverage for higher pay. Maybe a vendor exists that will contract out support, but then there are those costs to consider as well. And let’s not forget end-user training – in our example, anyone that runs reports will need to be trained on how to use the new system. Here’s the Point We still tend to look at software in an “off the shelf” kind of way. It’s very easy to say “oh, this product is better than vendor x’s product – and its free because its OSS!” but the reality is that implementing any new technology within an organization has a cost regardless of the retail price of the product. Training, integration, support – these are real costs that impact an organization and span multiple departments. Whether you’re pitching an improved business process, a new system, or a new technology, you need to consider the bigger picture costs of implementation. What you define as success (in our example, having better reporting functionality) might not be what others define as success if implementing your solution causes them issues. A true enterprise solution needs to consider the entire enterprise.

  Read the article

• Ubuntu 12.04 + Wifi not working

  - by user171154

  i'm having problems connecting over wireless. At the moment, I'm using wicd. It seems to get stuck on "Verifying AP association...". Without wicd I can get the connection up and ping the Net - but if I take eth0 down (ifconfig eth0 down), my wireless goes away too (same result if I unplug the wire instead). wicd is the only way I can bring eth0 back (which is the main reason I'm using it) - ifconfig eth0 and/or ifup eth0 do not re-enable the connection (I just discovered it leaves out the gateway. Adding the gateway back in re-enables the connection including wifi; I didn't want to delete the info about wicd above in case it gives someone an idea.) Doing it manually, despite the errors (which it would be nice to also resolve) - allows me to ping the outside world: ifup wlan0 ioctl[SIOCSIWENCODEEXT]: Invalid argument ioctl[SIOCSIWENCODEEXT]: Invalid argument ssh stop/waiting ssh start/running, process 17336 ping -I wlan0 -c 4 8.8.8.8 PING 8.8.8.8 (8.8.8.8) from 192.168.0.12 wlan0: 56(84) bytes of data. 64 bytes from 8.8.8.8: icmp_req=1 ttl=43 time=48.8 ms 64 bytes from 8.8.8.8: icmp_req=2 ttl=43 time=47.9 ms 64 bytes from 8.8.8.8: icmp_req=3 ttl=43 time=48.7 ms 64 bytes from 8.8.8.8: icmp_req=4 ttl=43 time=53.2 ms --- 8.8.8.8 ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3003ms rtt min/avg/max/mdev = 47.975/49.711/53.235/2.063 ms # iwconfig lo no wireless extensions. wlan0 IEEE 802.11bgn ESSID:"TPLINK" Mode:Managed Frequency:2.427 GHz Access Point: 64:66:xx:xx:xx:22 Bit Rate=108 Mb/s Tx-Power=27 dBm Retry long limit:7 RTS thr:off Fragment thr:off Encryption key:off Power Management:off Link Quality=70/70 Signal level=-39 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:3 Missed beacon:0 bus info: pci@0000:03:00.0 logical name: wlan0 version: 01 serial: f0:7d:68:c1:b4:13 width: 64 bits clock: 33MHz capabilities: pm msi pciexpress msix bus_master cap_list ethernet physical wireless configuration: broadcast=yes driver=ath9k driverversion=3.2.0-67-generic-pae firmware=N/A latency=0 link=no multicast=yes wireless=IEEE 802.11bgn resources: irq:17 memory:dfbf0000-dfbfffff ip route default via 192.168.0.1 dev eth0 default via 192.168.0.1 dev wlan0 metric 100 169.254.0.0/16 dev wlan0 scope link metric 1000 192.168.0.0/24 dev eth0 proto kernel scope link src 192.168.0.102 192.168.0.0/24 dev wlan0 proto kernel scope link src 192.168.0.12 (For the record, I have no idea what the 169.254.0.0 address is doing there.) uname -a 3.2.0-67-generic-pae #101-Ubuntu SMP Tue Jul 15 18:04:54 UTC 2014 i686 i686 i386 GNU/Linux lshw -C network *-network description: Ethernet interface product: NetXtreme BCM5751 Gigabit Ethernet PCI Express vendor: Broadcom Corporation physical id: 0 bus info: pci@0000:02:00.0 logical name: eth0 version: 01 serial: 00:11:11:59:fc:09 size: 100Mbit/s capacity: 1Gbit/s width: 64 bits clock: 33MHz capabilities: pm vpd msi pciexpress bus_master cap_list ethernet physical tp 10bt 10bt-fd 100bt 100bt-fd 1000bt 1000bt-fd autonegotiation configuration: autonegotiation=on broadcast=yes driver=tg3 driverversion=3.121 duplex=full firmware=5751-v3.23a ip=192.168.0.102 latency=0 link=yes multicast=yes port=twisted pair speed=100Mbit/s resources: irq:16 memory:dfcf0000-dfcfffff *-network description: Wireless interface product: AR5418 Wireless Network Adapter [AR5008E 802.11(a)bgn] (PCI-Express) vendor: Qualcomm Atheros physical id: 0 /etc/network/interfaces # interfaces(5) file used by ifup(8) and ifdown(8) auto lo iface lo inet loopback source /etc/network/interfaces.eth0 source /etc/network/interfaces.wlan0 /etc/network/interfaces.eth0 #Main Interface auto eth0 iface eth0 inet static address 192.168.0.102 netmask 255.255.255.0 gateway 192.168.0.1 /etc/network/interfaces.wlan0 auto wlan0 iface wlan0 inet static address 192.168.0.12 gateway 192.168.0.1 dns-nameservers 192.168.0.1 8.8.8.8 netmask 255.255.255.0 wpa-driver wext wpa-ssid TPLINK wpa-ap-scan 1 wpa-proto RSN wpa-pairwise CCMP wpa-group CCMP wpa-key-mgmt WPA-PSK wpa-psk dca1badb5fd4e9axxx4xxdaaxxfa91xx610bxx6a7d57ef67af9809dxx6af42e39 /etc/wpa_supplicant.conf ctrl_interface=/var/run/wpa_supplicant network={ ssid="TPLINK" psk="my password" key_mgmt=WPA-PSK proto=RSN pairwise=CCMP group=CCMP } ifdown eth0 ifdown: interface eth0 not configured ifconfig eth0 Link encap:Ethernet HWaddr 00:11:xx:xx:xx:09 inet addr:192.168.0.102 Bcast:192.168.0.255 Mask:255.255.255.0 inet6 addr: fe80::211:11ff:fe59:fc09/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:213690 errors:0 dropped:0 overruns:0 frame:0 TX packets:155266 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:220057808 (220.0 MB) TX bytes:21137696 (21.1 MB) Interrupt:16 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:196412 errors:0 dropped:0 overruns:0 frame:0 TX packets:196412 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:153270697 (153.2 MB) TX bytes:153270697 (153.2 MB) wlan0 Link encap:Ethernet HWaddr f0:7d:xx:xx:xx:13 inet addr:192.168.0.12 Bcast:192.168.0.255 Mask:255.255.255.0 inet6 addr: fe80::f27d:68ff:fec1:b413/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:11335 errors:0 dropped:0 overruns:0 frame:0 TX packets:7287 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:2563290 (2.5 MB) TX bytes:855746 (855.7 KB) ifconfig eth0 down ifconfig eth0 Link encap:Ethernet HWaddr 00:xx:xx:xx:xx:09 inet addr:192.168.0.102 Bcast:192.168.0.255 Mask:255.255.255.0 inet6 addr: fe80::211:11ff:fe59:fc09/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:2 errors:0 dropped:0 overruns:0 frame:0 TX packets:1 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:192 (192.0 B) TX bytes:94 (94.0 B) Interrupt:16 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:196418 errors:0 dropped:0 overruns:0 frame:0 TX packets:196418 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:153270871 (153.2 MB) TX bytes:153270871 (153.2 MB) wlan0 Link encap:Ethernet HWaddr f0:7d:xx:xx:xx:13 inet addr:192.168.0.12 Bcast:192.168.0.255 Mask:255.255.255.0 inet6 addr: fe80::f27d:68ff:fec1:b413/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:11359 errors:0 dropped:0 overruns:0 frame:0 TX packets:7293 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:2565482 (2.5 MB) TX bytes:856363 (856.3 KB) ip route default via 192.168.0.1 dev wlan0 metric 100 169.254.0.0/16 dev wlan0 scope link metric 1000 192.168.0.0/24 dev wlan0 proto kernel scope link src 192.168.0.12 192.168.0.0/24 dev eth0 proto kernel scope link src 192.168.0.102 ping -I wlan0 -c 4 8.8.8.8 PING 8.8.8.8 (8.8.8.8) from 192.168.0.12 wlan0: 56(84) bytes of data. --- 8.8.8.8 ping statistics --- 4 packets transmitted, 0 received, 100% packet loss, time 3024ms ping -I eth0 -c 3 router PING router (192.168.0.1) from 192.168.0.102 eth0: 56(84) bytes of data. --- router ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2015ms ping -I wlan0 -c 3 router PING router (192.168.0.1) from 192.168.0.12 wlan0: 56(84) bytes of data. --- router ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2014ms Let me know if you need more info. Thank you in advance.

  Read the article

• Projected Results

  - by Sylvie MacKenzie, PMP

  Excerpt from PROFIT - ORACLE - by Monica Mehta Yasser Mahmud has seen a revolution in project management over the past decade. During that time, the former Primavera product strategist (who joined Oracle when his company was acquired in 2008) has not only observed a transformation in the way IT systems support corporate projects but the role project portfolio management (PPM) plays in the enterprise. “15 years ago project management was the domain of project management office (PMO),” Mahmud recalls of earlier days. “But over the course of the past decade, we've seen it transform into a mission critical enterprise discipline, that has made Primavera indispensable in the board room. Now, as a senior manager, a board member, or a C-level executive you have direct and complete visibility into what’s kind of going on in the organization—at a level of detail that you're going to consume that information.” Now serving as Oracle’s vice president of product strategy and industry marketing, Mahmud shares his thoughts on how Oracle’s Primavera solutions have evolved and how best-in-class project portfolio management systems can help businesses stay competitive. Profit: What do you feel are the market dynamics that are changing project management today? Mahmud: First, the data explosion. We're generating data at twice the rate at which we can actually store it. The same concept applies for project-intensive organizations. A lot of data is gathered, but what are we really doing with it? Are we turning data into insight? Are we using that insight and turning it into foresight with analytics tools? This is a key driver that will separate the very good companies—the very competitive companies—from those that are not as competitive. Another trend is centered on the explosion of mobile computing. By the year 2013, an estimated 35 percent of the world’s workforce is going to be mobile. That’s one billion people. So the question is not if you're going to go mobile, it’s how fast you are going to go mobile. What kind of impact does that have on how the workforce participates in projects? What worked ten to fifteen years ago is not going to work today. It requires a real rethink around the interfaces and how data is actually presented. Profit: What is the role of project management in this new landscape? Mahmud: We recently conducted a PPM study with the Economist Intelligence Unit centered to determine how important project management is considered within organizations. Our target was primarily CFOs, CIOs, and senior managers and we discovered that while 95 percent of participants believed it critical to their business, only six percent were confident that projects were delivered on time and on budget. That’s a huge gap. Most organizations are looking for efficiency, especially in these volatile financial times. But senior management can’t keep track of every project in a large organization. As a result, executives are attempting to inventory the work being conducted under their watch. What is often needed is a very high-level assessment conducted at the board level to say, “Here are the 50 initiatives that we have underway. How do they line up with our strategic drivers?” This line of questioning can provide early warning that work and strategy are out of alignment; finding the gap between what the business needs to do and the actual performance scorecard. That’s low-hanging fruit for any executive looking to increase efficiency and save money. But it can only be obtained through proper assessment of existing projects—and you need a project system of record to get that done. Over the next decade or so, project management is going to transform into holistic work management. Business leaders will want make sure key projects align with corporate strategy, but also the ability to drill down into daily activity and smaller projects to make sure they line up as well. Keeping employees from working on tasks—even for a few hours—that don’t line up with corporate goals will, in many ways, become a competitive differentiator. Profit: How do all of these market challenges and shifting trends impact Oracle’s Primavera solutions and meeting customers’ needs? Mahmud: For Primavera, it’s a transformation from being a project management application to a PPM system in the enterprise. Also making that system a mission-critical application by connecting to other key applications within the ecosystem, such as the enterprise resource planning (ERP), supply chain, and CRM systems. Analytics have also become a huge component. Business analytics have made Oracle’s Primavera applications pertinent in the boardroom. Now, as a senior manager, a board member, a CXO, CIO, or CEO, you have direct visibility into what’s going on in the organization at a level that you're able to consume that information. In addition, all of this information pairs up really well with your financials and other data. Certainly, when you're an Oracle shop, you have that visibility that you didn’t have before from a project execution perspective. Profit: What new strategies and tools are being implemented to create a more efficient workplace for users? Mahmud: We believe very strongly that just because you call something an enterprise project portfolio management system doesn’t make it so—you have to get people to want to participate in the system. This can’t be mandated down from the top. It simply doesn’t work that way. A truly adoptable solution is one that makes it super easy for all types users to participate, by providing them interfaces where they live. Keeping that in mind, a major area of development has been alternative user interfaces. This is increasingly resulting in the creation of lighter weight, targeted interfaces such as iOS applications, and smartphones interfaces such as for iPhone and Android platform. Profit: How does this translate into the development of Oracle’s Primavera solutions? Mahmud: Let me give you a few examples. We recently announced the launch of our Primavera P6 Team Member application, which is a native iOS application for the iPhone. This interface makes it easier for team members to do their jobs quickly and effectively. Similarly, we introduced the Primavera analytics application, which can be consumed via mobile devices, and when married with Oracle Spatial capabilities, users can get a geographical view of what’s going on and which projects are occurring in various locations around the world. Lastly, we introduced advanced email integration that allows project team members to status work via E-mail. This functionality leverages the fact that users are in E-mail system throughout the day and allows them to status their work without the need to launch the Primavera application. It comes back to a mantra: provide as many alternative user interfaces as possible, so you can give people the ability to work, to participate, to raise issues, to create projects, in the places where they live. Do it in such a way that it’s non-intrusive, do it in such a way that it’s easy and intuitive and they can get it done in a short amount of time. If you do that, workers can get back to doing what they're actually getting paid for.

  Read the article

• Character jump animation is not working when i hit the space bar

  - by muzzy

  i am having an issue with my game in XNA. My jump sprite sheet for my character does not trigger when i hit the space bar. I cant seem to find the problem. Please help me. I am also put the code below to make things easier. namespace WindowsGame4 { public class Game1 : Microsoft.Xna.Framework.Game { GraphicsDeviceManager graphics; SpriteBatch spriteBatch; // start of new code Texture2D playerWalk; // sprite sheet of walk cycle (14 frames) Texture2D idle; // idle animation Texture2D jump; // jump animation Vector2 playerPos; // to hold x and y position info for the player Point frameDimensions; // to hold width and height values for the frames int presentFrame; // to record which frame we are on at any given time int noOfFrames; // to hold the total number of frames in the spritesheet int elapsedTime; // to know how long each frame has been shown int frameDuration; // to hold info about how long each frame should be shown SpriteEffects flipDirection; // SpriteEffects object int speed; //rate of movement int upMovement; int downMovement; int rightMovement; int leftMovement; int jumpApex; string state; //this is going to be "idle","walking" or "jumping". KeyboardState previousKeyboardState; Vector2 originalPlayerPos; Vector2 movementDirection; Vector2 movementSpeed; public Game1() { graphics = new GraphicsDeviceManager(this); Content.RootDirectory = "Content"; } protected override void Initialize() { // textures will be defined in the LoadContent() method playerPos = new Vector2(0, 200); // starting position for the player is at the left of the screen, and a Y position of 200 frameDimensions = new Point(55, 65); // each frame in the idle sprite sheet is 55 wide by 65 high presentFrame = 0; // start at frame 0 noOfFrames = 5; // there are 5 frames in the idle cycle elapsedTime = 0; // set elapsed time to start at 0 frameDuration = 80; // 80 milliseconds is how long each frame will show for (the higher the number, the slower the animation) flipDirection = SpriteEffects.None; // set the value of flipDirection to none speed = 200; upMovement = -2; downMovement = 2; rightMovement = 1; leftMovement = -1; jumpApex = 100; state = "idle"; previousKeyboardState = Keyboard.GetState(); originalPlayerPos = Vector2.Zero; movementDirection = Vector2.Zero; movementSpeed = Vector2.Zero; base.Initialize(); } protected override void LoadContent() { spriteBatch = new SpriteBatch(GraphicsDevice); playerWalk = Content.Load<Texture2D>("sprites/walkSmall"); // load the walk cycle spritesheet idle = Content.Load<Texture2D>("sprites/idleCycle"); // load the idle cycle sprite sheet jump = Content.Load<Texture2D>("sprites/jump"); // load the jump cycle sprite sheet } protected override void UnloadContent() // we're not using this method at the moment { } protected override void Update(GameTime gameTime) // Update method - used it to call a number of other methods { if (Keyboard.GetState().IsKeyDown(Keys.Escape)) { this.Exit(); // Exit the game if the Escape key is pressed } KeyboardState presentKeyboardState = Keyboard.GetState(); UpdateMovement(presentKeyboardState, gameTime); UpdateIdle(presentKeyboardState, gameTime); UpdateJump(presentKeyboardState); UpdateAnimation(gameTime); playerPos += movementDirection * movementSpeed * (float)gameTime.ElapsedGameTime.TotalSeconds; previousKeyboardState = presentKeyboardState; base.Update(gameTime); } private void UpdateAnimation(GameTime gameTime) { elapsedTime += gameTime.ElapsedGameTime.Milliseconds; if (elapsedTime > frameDuration) { elapsedTime -= frameDuration; elapsedTime = elapsedTime - frameDuration; presentFrame++; if (presentFrame > noOfFrames) if (state != "jumping") { presentFrame = 0; } else { presentFrame = 8; } } } protected void UpdateMovement(KeyboardState presentKeyboardState, GameTime gameTime) { if (state == "idle") { movementSpeed = Vector2.Zero; movementDirection = Vector2.Zero; if (presentKeyboardState.IsKeyDown(Keys.Left)) { state = "walking"; movementSpeed.X = speed; movementDirection.X = leftMovement; flipDirection = SpriteEffects.FlipHorizontally; } if (presentKeyboardState.IsKeyDown(Keys.Right)) { state = "walking"; movementSpeed.X = speed; movementDirection.X = rightMovement; flipDirection = SpriteEffects.None; } } } private void UpdateIdle(KeyboardState presentKeyboardState, GameTime gameTime) { if ((presentKeyboardState.IsKeyUp(Keys.Left) && previousKeyboardState.IsKeyDown(Keys.Left) || presentKeyboardState.IsKeyUp(Keys.Right) && previousKeyboardState.IsKeyDown(Keys.Right) && state != "jumping")) { state = "idle"; } } private void UpdateJump(KeyboardState presentKeyboardState) { if (state == "walking" || state == "idle") { if (presentKeyboardState.IsKeyDown(Keys.Space) && !presentKeyboardState.IsKeyDown(Keys.Space)) { presentFrame = 1; DoJump(); } } if (state == "jumping") { if (originalPlayerPos.Y - playerPos.Y > jumpApex) { movementDirection.Y = downMovement; } if (playerPos.Y > originalPlayerPos.Y) { playerPos.Y = originalPlayerPos.Y; state = "idle"; movementDirection = Vector2.Zero; } } } private void DoJump() { if (state != "jumping") { state = "jumping"; originalPlayerPos = playerPos; movementDirection.Y = upMovement; movementSpeed = new Vector2(speed, speed); } } protected override void Draw(GameTime gameTime) // Draw method { GraphicsDevice.Clear(Color.CornflowerBlue); spriteBatch.Begin(); // begin the spritebatch if (state == "walking") { noOfFrames = 14; frameDimensions = new Point(55, 65); Vector2 playerWalkPos = new Vector2(playerPos.X, playerPos.Y - 28); spriteBatch.Draw(playerWalk, playerWalkPos, new Rectangle((presentFrame * frameDimensions.X), 0, frameDimensions.X, frameDimensions.Y), Color.White, 0, Vector2.Zero, 1, flipDirection, 0); } if (state == "idle") { noOfFrames = 5; frameDimensions = new Point(55, 65); Vector2 idlePos = new Vector2(playerPos.X, playerPos.Y - 28); spriteBatch.Draw(idle, idlePos, new Rectangle((presentFrame * frameDimensions.X), 0, frameDimensions.X, frameDimensions.Y), Color.White, 0, Vector2.Zero, 1, flipDirection, 0); } if (state == "jumping") { noOfFrames = 9; frameDimensions = new Point(55, 92); Vector2 jumpPos = new Vector2(playerPos.X, playerPos.Y - 28); spriteBatch.Draw(jump, jumpPos, new Rectangle((presentFrame * frameDimensions.X), 0, frameDimensions.X, frameDimensions.Y), Color.White, 0, Vector2.Zero, 1, flipDirection, 0); } spriteBatch.End(); // end the spritebatch commands base.Draw(gameTime); } } }

  Read the article

• Organization &amp; Architecture UNISA Studies &ndash; Chap 4

  - by MarkPearl

  Learning Outcomes Explain the characteristics of memory systems Describe the memory hierarchy Discuss cache memory principles Discuss issues relevant to cache design Describe the cache organization of the Pentium Computer Memory Systems There are key characteristics of memory… Location – internal or external Capacity – expressed in terms of bytes Unit of Transfer – the number of bits read out of or written into memory at a time Access Method – sequential, direct, random or associative From a users perspective the two most important characteristics of memory are… Capacity Performance – access time, memory cycle time, transfer rate The trade off for memory happens along three axis… Faster access time, greater cost per bit Greater capacity, smaller cost per bit Greater capacity, slower access time This leads to people using a tiered approach in their use of memory   As one goes down the hierarchy, the following occurs… Decreasing cost per bit Increasing capacity Increasing access time Decreasing frequency of access of the memory by the processor The use of two levels of memory to reduce average access time works in principle, but only if conditions 1 to 4 apply. A variety of technologies exist that allow us to accomplish this. Thus it is possible to organize data across the hierarchy such that the percentage of accesses to each successively lower level is substantially less than that of the level above. A portion of main memory can be used as a buffer to hold data temporarily that is to be read out to disk. This is sometimes referred to as a disk cache and improves performance in two ways… Disk writes are clustered. Instead of many small transfers of data, we have a few large transfers of data. This improves disk performance and minimizes processor involvement. Some data designed for write-out may be referenced by a program before the next dump to disk. In that case the data is retrieved rapidly from the software cache rather than slowly from disk. Cache Memory Principles Cache memory is substantially faster than main memory. A caching system works as follows.. When a processor attempts to read a word of memory, a check is made to see if this in in cache memory… If it is, the data is supplied, If it is not in the cache, a block of main memory, consisting of a fixed number of words is loaded to the cache. Because of the phenomenon of locality of references, when a block of data is fetched into the cache, it is likely that there will be future references to that same memory location or to other words in the block. Elements of Cache Design While there are a large number of cache implementations, there are a few basic design elements that serve to classify and differentiate cache architectures… Cache Addresses Cache Size Mapping Function Replacement Algorithm Write Policy Line Size Number of Caches Cache Addresses Almost all non-embedded processors support virtual memory. Virtual memory in essence allows a program to address memory from a logical point of view without needing to worry about the amount of physical memory available. When virtual addresses are used the designer may choose to place the cache between the MMU (memory management unit) and the processor or between the MMU and main memory. The disadvantage of virtual memory is that most virtual memory systems supply each application with the same virtual memory address space (each application sees virtual memory starting at memory address 0), which means the cache memory must be completely flushed with each application context switch or extra bits must be added to each line of the cache to identify which virtual address space the address refers to. Cache Size We would like the size of the cache to be small enough so that the overall average cost per bit is close to that of main memory alone and large enough so that the overall average access time is close to that of the cache alone. Also, larger caches are slightly slower than smaller ones. Mapping Function Because there are fewer cache lines than main memory blocks, an algorithm is needed for mapping main memory blocks into cache lines. The choice of mapping function dictates how the cache is organized. Three techniques can be used… Direct – simplest technique, maps each block of main memory into only one possible cache line Associative – Each main memory block to be loaded into any line of the cache Set Associative – exhibits the strengths of both the direct and associative approaches while reducing their disadvantages For detailed explanations of each approach – read the text book (page 148 – 154) Replacement Algorithm For associative and set associating mapping a replacement algorithm is needed to determine which of the existing blocks in the cache must be replaced by a new block. There are four common approaches… LRU (Least recently used) FIFO (First in first out) LFU (Least frequently used) Random selection Write Policy When a block resident in the cache is to be replaced, there are two cases to consider If no writes to that block have happened in the cache – discard it If a write has occurred, a process needs to be initiated where the changes in the cache are propagated back to the main memory. There are several approaches to achieve this including… Write Through – all writes to the cache are done to the main memory as well at the point of the change Write Back – when a block is replaced, all dirty bits are written back to main memory The problem is complicated when we have multiple caches, there are techniques to accommodate for this but I have not summarized them. Line Size When a block of data is retrieved and placed in the cache, not only the desired word but also some number of adjacent words are retrieved. As the block size increases from very small to larger sizes, the hit ratio will at first increase because of the principle of locality, which states that the data in the vicinity of a referenced word are likely to be referenced in the near future. As the block size increases, more useful data are brought into cache. The hit ratio will begin to decrease as the block becomes even bigger and the probability of using the newly fetched information becomes less than the probability of using the newly fetched information that has to be replaced. Two specific effects come into play… Larger blocks reduce the number of blocks that fit into a cache. Because each block fetch overwrites older cache contents, a small number of blocks results in data being overwritten shortly after they are fetched. As a block becomes larger, each additional word is farther from the requested word and therefore less likely to be needed in the near future. The relationship between block size and hit ratio is complex, and no set approach is judged to be the best in all circumstances.   Pentium 4 and ARM cache organizations The processor core consists of four major components: Fetch/decode unit – fetches program instruction in order from the L2 cache, decodes these into a series of micro-operations, and stores the results in the L2 instruction cache Out-of-order execution logic – Schedules execution of the micro-operations subject to data dependencies and resource availability – thus micro-operations may be scheduled for execution in a different order than they were fetched from the instruction stream. As time permits, this unit schedules speculative execution of micro-operations that may be required in the future Execution units – These units execute micro-operations, fetching the required data from the L1 data cache and temporarily storing results in registers Memory subsystem – This unit includes the L2 and L3 caches and the system bus, which is used to access main memory when the L1 and L2 caches have a cache miss and to access the system I/O resources

  Read the article

• Unable to center text in IE but works in firefox

  - by greenpool

  Can somebody point out where I'm going wrong with the following code. Text inside td elements need to be centered except for Summary and Experience. This only appears to work in Firefox/chrome. In IE8 all td text are displayed as left-justified. No matter what I try it doesn't center it. Any particular reason why this would happen? Thanks. css #viewAll { font-family:"Trebuchet MS", Arial, Helvetica, sans-serif; width:100%; border-collapse:collapse; margin-left:10px; table-layout: fixed; } #viewAll td, #viewAll th { font-size:1.1em; border:1px solid #98bf21; word-wrap:break-word; text-align:center; overflow:hidden; } #viewAll tbody td{ padding:2px; } #viewAll th { font-size:1.1em; padding-top:5px; padding-bottom:4px; background-color:#A7C942; color:#ffffff; } table <?php echo '<table id="viewAll" class="tablesorter">'; echo '<thead>'; echo '<tr align="center">'; echo '<th style="width:70px;">Product</th>'; echo '<th style="width:105px;">Prob</th>'; echo '<th style="width:105px;">I</th>'; echo '<th style="width:60px;">Status</th>'; echo '<th style="width:120px;">Experience</th>'; echo '<th style="width:200px;">Technical Summary</th>'; echo '<th style="width:80px;">Record Created</th>'; echo '<th style="width:80px;">Record Updated</th>'; echo '<th style="width:50px;">Open</th>'; echo '</tr>'; echo '</thead>'; echo '<tbody>'; while ($data=mysqli_fetch_array($result)){ #limiting the summary text displayed in the table $limited_summary = (strlen($data['summary']) > 300) ? substr(($data['summary']),0,300) . '...' : $data['summary']; $limited_exp = (strlen($data['exp']) > 300) ? substr(($data['exp']),0,300) . '...' : $data['exp']; echo '<tr align="center"> <td style="width:70px; text-align:center;">'.$data['product'].'</td>'; //if value is '-' do not display as link if ($data['prob'] != '-'){ echo '<td style="width:105px;">'.$data['prob'].'</a></td>'; } else{ echo '<td style="width:105px; ">'.$data['prob'].'</td>'; } if ($data['i'] != '-'){ echo '<td style="width:105px; ">'.$data['i'].'</a></td>'; } else{ echo '<td style="width:105px; ">'.$data['i'].'</td>'; } echo'<td style="width:40px; " >'.$data['status'].'</td> <td style="width:120px; text-align:left;">'.$limited_cust_exp.'</td> <td style="width:200px; text-align:left;">'.$limited_summary.'</td> <td style="width:80px; ">'.$data['created'].'</td> <td style="width:80px; ">'.$data['updated'].'</td>'; if (isset($_SESSION['username'])){ echo '<td style="width:50px; "> <form action="displayRecord.php" method="get">'.' <input type="hidden" name="id" value="'. $data['id'].'" style="text-decoration: none" /><input type="submit" value="Open" /></form></td>'; }else{ echo '<td style="width:50px; "> <form action="displayRecord.php" method="get">'.' <input type="hidden" name="id" value="'. $data['id'].'" style="text-decoration: none" /><input type="submit" value="View" /></form></td>'; } echo '</tr>'; }#end of while echo '</tbody>'; echo '</table>'; ?>

  Read the article

• ColdFusion Server crash after thousands of HTTP requests

  - by Jason Bristol

  We are running ColdFusion 8 on a windows server 2003 VPS with an API that exposes student records to a partner API through a connector. Our API returns around 50k student records serialized in XML format pretty seamlessly. My question originates when something very frightening happened today when we tested our connector to our partners API. Our entire website and web host went down. We assumed that our host was just having some issues and after 4 hours with no resolution and no response from their customer service we finally got a response from them claiming that they had an "unauthorized user" in their network. After our server was back up we were unable to connect to our website as if the web service or coldfusion itself had froze. This is really where my concern comes from as I fear we may have overloaded the web service. As I mentioned before we tried sending over 50k HTTP POST requests over to our partner's API, however everything stopped after around 1.6k Is this bad practice or is there some sort of rate limiting I can relax somewhere in server configuration? We managed to find a workaround, but it bypasses our connector which is critical to our design. This would have been a one time deal as the purpose of so many requests was to populate our partner's website with current data, after that hourly syncs will keep requests down to around 100 per hour. UPDATE Our partner API is owned and operated by Pardot. We are converting students to prospects by passing student data to their API which unfortunately only seems to accept one student at a time. For that reason we have to do all 50k requests individually. Our server has 4GB of RAM, an Intel Core 2 Duo @ 2.8GHz running Windows Server 2003 SP2. I monitored the server during a 100 student sync, a 400 student sync, and a 1.4k student sync with the following results: 100 students - 2.25GB of Memory, 30-40% CPU utilization, 0.2-0.3% network bandwidth 400 students - 2.30GB of Memory, 30-50% CPU utilization, 0.2-1.0% network bandwidth 1.4k students - 2.30GB of Memory, 30-70% CPU utilization, 0.2-1.0% network bandwidth I know this is a far cry from 50k students, but I don't want to risk taking down our CMS system again assuming that was the cause. To give you a look at our code: <cfif (#getStudents.statusCode# eq "200 OK")> <cftry> <cfloop index="StudentXML" array="#XmlSearch(responseSTUD,'/students/student')#"> <cfset StudentXML = XmlParse(StudentXML)> <cfhttp url="#PARDOT_CMS_UPSERT#" method="post" timeout="10000" > <cfhttpparam type="url" name="user_key" value="#PARDOT_CMS_USERKEY#"> <cfhttpparam type="url" name="api_key" value="#api_key#"> <cfhttpparam type="url" name="email" value="#StudentXML.student.email.XmlText#"> <cfhttpparam type="url" name="first_name" value="#StudentXML.student.first.XmlText#"> <cfhttpparam type="url" name="last_name" value="#StudentXML.student.last.XmlText#"> <cfhttpparam type="url" name="in_cms" value="#StudentXML.student.studentid.XmlText#"> <cfhttpparam type="url" name="company" value="#StudentXML.student.agencyname.XmlText#"> <cfhttpparam type="url" name="country" value="#StudentXML.student.countryname.XmlText#"> <cfhttpparam type="url" name="address_one" value="#StudentXML.student.address.XmlText#"> <cfhttpparam type="url" name="address_two" value="#StudentXML.student.address2.XmlText#"> <cfhttpparam type="url" name="city" value="#StudentXML.student.city.XmlText#"> <cfhttpparam type="url" name="state" value="#StudentXML.student.state_province.XmlText#"> <cfhttpparam type="url" name="zip" value="#StudentXML.student.postalcode.XmlText#"> <cfhttpparam type="url" name="phone" value="#StudentXML.student.phone.XmlText#"> <cfhttpparam type="url" name="fax" value="#StudentXML.student.fax.XmlText#"> <cfhttpparam type="url" name="output" value="simple"> </cfhttp> </cfloop> <cfcatch type="any"> <cfdump var="#cfcatch.Message#"> </cfcatch> </cftry> </cfif> UPDATE 2 I checked the CF logs and found a couple of these: "Error","jrpp-8","06/06/13","16:10:18","CMS-API","Java heap space The specific sequence of files included or processed is: D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm, line: 675 " java.lang.OutOfMemoryError: Java heap space at java.util.Arrays.copyOf(Arrays.java:2882) at java.io.CharArrayWriter.write(CharArrayWriter.java:105) at coldfusion.runtime.CharBuffer.replace(CharBuffer.java:37) at coldfusion.runtime.CharBuffer.replace(CharBuffer.java:50) at coldfusion.runtime.NeoBodyContent.write(NeoBodyContent.java:254) at cfapi2ecfm292155732._factor30(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:675) at cfapi2ecfm292155732._factor31(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:662) at cfapi2ecfm292155732._factor36(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:659) at cfapi2ecfm292155732._factor42(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:657) at cfapi2ecfm292155732._factor37(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm) at cfapi2ecfm292155732._factor44(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:456) at cfapi2ecfm292155732._factor38(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm) at cfapi2ecfm292155732._factor46(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:455) at cfapi2ecfm292155732._factor39(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm) at cfapi2ecfm292155732._factor47(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:453) at cfapi2ecfm292155732.runPage(D:\Clients\www.xxx.com\www\dev.cms\api\v1\api.cfm:1) at coldfusion.runtime.CfJspPage.invoke(CfJspPage.java:192) at coldfusion.tagext.lang.IncludeTag.doStartTag(IncludeTag.java:366) at coldfusion.filter.CfincludeFilter.invoke(CfincludeFilter.java:65) at coldfusion.filter.ApplicationFilter.invoke(ApplicationFilter.java:279) at coldfusion.filter.RequestMonitorFilter.invoke(RequestMonitorFilter.java:48) at coldfusion.filter.MonitoringFilter.invoke(MonitoringFilter.java:40) at coldfusion.filter.PathFilter.invoke(PathFilter.java:86) at coldfusion.filter.ExceptionFilter.invoke(ExceptionFilter.java:70) at coldfusion.filter.ClientScopePersistenceFilter.invoke(ClientScopePersistenceFilter.java:28) at coldfusion.filter.BrowserFilter.invoke(BrowserFilter.java:38) at coldfusion.filter.NoCacheFilter.invoke(NoCacheFilter.java:46) at coldfusion.filter.GlobalsFilter.invoke(GlobalsFilter.java:38) at coldfusion.filter.DatasourceFilter.invoke(DatasourceFilter.java:22) at coldfusion.CfmServlet.service(CfmServlet.java:175) at coldfusion.bootstrap.BootstrapServlet.service(BootstrapServlet.java:89) at jrun.servlet.FilterChain.doFilter(FilterChain.java:86) Looks like I might have crashed the JVM in CF, is there a better way to do this? We are thinking of just exporting all records initially as a CSV file and importing it into Pardot seeing as we will never have to do a request this large again.

  Read the article

• Troubleshooting latency spikes on ESXi NFS datastores

  - by exo_cw

  I'm experiencing fsync latencies of around five seconds on NFS datastores in ESXi, triggered by certain VMs. I suspect this might be caused by VMs using NCQ/TCQ, as this does not happen with virtual IDE drives. This can be reproduced using fsync-tester (by Ted Ts'o) and ioping. For example using a Grml live system with a 8GB disk: Linux 2.6.33-grml64: root@dynip211 /mnt/sda # ./fsync-tester fsync time: 5.0391 fsync time: 5.0438 fsync time: 5.0300 fsync time: 0.0231 fsync time: 0.0243 fsync time: 5.0382 fsync time: 5.0400 [... goes on like this ...] That is 5 seconds, not milliseconds. This is even creating IO-latencies on a different VM running on the same host and datastore: root@grml /mnt/sda/ioping-0.5 # ./ioping -i 0.3 -p 20 . 4096 bytes from . (reiserfs /dev/sda): request=1 time=7.2 ms 4096 bytes from . (reiserfs /dev/sda): request=2 time=0.9 ms 4096 bytes from . (reiserfs /dev/sda): request=3 time=0.9 ms 4096 bytes from . (reiserfs /dev/sda): request=4 time=0.9 ms 4096 bytes from . (reiserfs /dev/sda): request=5 time=4809.0 ms 4096 bytes from . (reiserfs /dev/sda): request=6 time=1.0 ms 4096 bytes from . (reiserfs /dev/sda): request=7 time=1.2 ms 4096 bytes from . (reiserfs /dev/sda): request=8 time=1.1 ms 4096 bytes from . (reiserfs /dev/sda): request=9 time=1.3 ms 4096 bytes from . (reiserfs /dev/sda): request=10 time=1.2 ms 4096 bytes from . (reiserfs /dev/sda): request=11 time=1.0 ms 4096 bytes from . (reiserfs /dev/sda): request=12 time=4950.0 ms When I move the first VM to local storage it looks perfectly normal: root@dynip211 /mnt/sda # ./fsync-tester fsync time: 0.0191 fsync time: 0.0201 fsync time: 0.0203 fsync time: 0.0206 fsync time: 0.0192 fsync time: 0.0231 fsync time: 0.0201 [... tried that for one hour: no spike ...] Things I've tried that made no difference: Tested several ESXi Builds: 381591, 348481, 260247 Tested on different hardware, different Intel and AMD boxes Tested with different NFS servers, all show the same behavior: OpenIndiana b147 (ZFS sync always or disabled: no difference) OpenIndiana b148 (ZFS sync always or disabled: no difference) Linux 2.6.32 (sync or async: no difference) It makes no difference if the NFS server is on the same machine (as a virtual storage appliance) or on a different host Guest OS tested, showing problems: Windows 7 64 Bit (using CrystalDiskMark, latency spikes happen mostly during preparing phase) Linux 2.6.32 (fsync-tester + ioping) Linux 2.6.38 (fsync-tester + ioping) I could not reproduce this problem on Linux 2.6.18 VMs. Another workaround is to use virtual IDE disks (vs SCSI/SAS), but that is limiting performance and the number of drives per VM. Update 2011-06-30: The latency spikes seem to happen more often if the application writes in multiple small blocks before fsync. For example fsync-tester does this (strace output): pwrite(3, "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"..., 1048576, 0) = 1048576 fsync(3) = 0 ioping does this while preparing the file: [lots of pwrites] pwrite(3, "********************************"..., 4096, 1036288) = 4096 pwrite(3, "********************************"..., 4096, 1040384) = 4096 pwrite(3, "********************************"..., 4096, 1044480) = 4096 fsync(3) = 0 The setup phase of ioping almost always hangs, while fsync-tester sometimes works fine. Is someone capable of updating fsync-tester to write multiple small blocks? My C skills suck ;) Update 2011-07-02: This problem does not occur with iSCSI. I tried this with the OpenIndiana COMSTAR iSCSI server. But iSCSI does not give you easy access to the VMDK files so you can move them between hosts with snapshots and rsync. Update 2011-07-06: This is part of a wireshark capture, captured by a third VM on the same vSwitch. This all happens on the same host, no physical network involved. I've started ioping around time 20. There were no packets sent until the five second delay was over: No. Time Source Destination Protocol Info 1082 16.164096 192.168.250.10 192.168.250.20 NFS V3 WRITE Call (Reply In 1085), FH:0x3eb56466 Offset:0 Len:84 FILE_SYNC 1083 16.164112 192.168.250.10 192.168.250.20 NFS V3 WRITE Call (Reply In 1086), FH:0x3eb56f66 Offset:0 Len:84 FILE_SYNC 1084 16.166060 192.168.250.20 192.168.250.10 TCP nfs > iclcnet-locate [ACK] Seq=445 Ack=1057 Win=32806 Len=0 TSV=432016 TSER=769110 1085 16.167678 192.168.250.20 192.168.250.10 NFS V3 WRITE Reply (Call In 1082) Len:84 FILE_SYNC 1086 16.168280 192.168.250.20 192.168.250.10 NFS V3 WRITE Reply (Call In 1083) Len:84 FILE_SYNC 1087 16.168417 192.168.250.10 192.168.250.20 TCP iclcnet-locate > nfs [ACK] Seq=1057 Ack=773 Win=4163 Len=0 TSV=769110 TSER=432016 1088 23.163028 192.168.250.10 192.168.250.20 NFS V3 GETATTR Call (Reply In 1089), FH:0x0bb04963 1089 23.164541 192.168.250.20 192.168.250.10 NFS V3 GETATTR Reply (Call In 1088) Directory mode:0777 uid:0 gid:0 1090 23.274252 192.168.250.10 192.168.250.20 TCP iclcnet-locate > nfs [ACK] Seq=1185 Ack=889 Win=4163 Len=0 TSV=769821 TSER=432716 1091 24.924188 192.168.250.10 192.168.250.20 RPC Continuation 1092 24.924210 192.168.250.10 192.168.250.20 RPC Continuation 1093 24.924216 192.168.250.10 192.168.250.20 RPC Continuation 1094 24.924225 192.168.250.10 192.168.250.20 RPC Continuation 1095 24.924555 192.168.250.20 192.168.250.10 TCP nfs > iclcnet_svinfo [ACK] Seq=6893 Ack=1118613 Win=32625 Len=0 TSV=432892 TSER=769986 1096 24.924626 192.168.250.10 192.168.250.20 RPC Continuation 1097 24.924635 192.168.250.10 192.168.250.20 RPC Continuation 1098 24.924643 192.168.250.10 192.168.250.20 RPC Continuation 1099 24.924649 192.168.250.10 192.168.250.20 RPC Continuation 1100 24.924653 192.168.250.10 192.168.250.20 RPC Continuation 2nd Update 2011-07-06: There seems to be some influence from TCP window sizes. I was not able to reproduce this problem using FreeNAS (based on FreeBSD) as a NFS server. The wireshark captures showed TCP window updates to 29127 bytes in regular intervals. I did not see them with OpenIndiana, which uses larger window sizes by default. I can no longer reproduce this problem if I set the following options in OpenIndiana and restart the NFS server: ndd -set /dev/tcp tcp_recv_hiwat 8192 # default is 128000 ndd -set /dev/tcp tcp_max_buf 1048575 # default is 1048576 But this kills performance: Writing from /dev/zero to a file with dd_rescue goes from 170MB/s to 80MB/s. Update 2011-07-07: I've uploaded this tcpdump capture (can be analyzed with wireshark). In this case 192.168.250.2 is the NFS server (OpenIndiana b148) and 192.168.250.10 is the ESXi host. Things I've tested during this capture: Started "ioping -w 5 -i 0.2 ." at time 30, 5 second hang in setup, completed at time 40. Started "ioping -w 5 -i 0.2 ." at time 60, 5 second hang in setup, completed at time 70. Started "fsync-tester" at time 90, with the following output, stopped at time 120: fsync time: 0.0248 fsync time: 5.0197 fsync time: 5.0287 fsync time: 5.0242 fsync time: 5.0225 fsync time: 0.0209 2nd Update 2011-07-07: Tested another NFS server VM, this time NexentaStor 3.0.5 community edition: Shows the same problems. Update 2011-07-31: I can also reproduce this problem on the new ESXi build 4.1.0.433742.

  Read the article

• Multiple routers, subnets, gateways etc

  - by allentown

  My current setup is: Cable modem dishes out 13 static IP's (/28), a GB switch is plugged into the cable modem, and has access to those 13 static IP's, I have about 6 "servers" in use right now. The cable modem is also a firewall, DHCP server, and 3 port 10/100 switch. I am using it as a firewall, but not currently as a DHCP server. I have plugged into the cable modem, two network cables, one which goes to the WAN port of a Linksys Dual Band Wireless 10/100/1000 router/switch. Into the linksys are a few workstations, a few printers, and some laptops connecting to wifi. I set the Linksys to use take static IP, and enabled DHCP for the workstations, printers, etc in 192.168.1.1/24. The network for the Linksys is mostly self contained, backups go to a SAN, on that network, it all happens through that switch, over GB. But I also get internet access from it as well via the cable modem using one static IP. This all works, however, I can not "see" the static IP machines when I am on the Linksys. I can get to them via ssh and other protocols, and if I want to from "outside", I open holes, like 80, 25, 587, 143, 22, etc. The second wire, from the cable modem/fireall/switch just uplinks to the managed GB switch. What are the pros and cons of this? I do not like giving up the static IP to the Linksys. I basically have a mixed network of public servers, and internal workstations. I want the public servers on public IP's because I do not want to mess with port forwarding and mappings. Is it correct also, that if someone breaches the Linksys wifi, they still would have a hard time getting to the static IP range, just by nature of the network topology? Today, just for a test, I toggled on the DHCP in the firewall/cable modem at 10.1.10.1/24 range, the Linksys is n the 192.168.1.100/24 range. At that point, all the static IP machines still had in and out access, but Linksys was unreachable. The cable modem only has 10/100 ports, so I will not plug anything but the network drop into it, which is 50Mb/10Mb. Which makes me think this could be less than ideal, as transfers from the workstation network to the server network will be bottlenecked at 100Mb when I have 1000Mb available. I may not need to solve that, if isolation is better though. I do not move a lot of data, if any, from Linsys network to server network, so for it to pretend to be remote is ok. Should I approach this any different? I could enable DHCP on the cable modem/firewall, it should still send out the statics to the GB switch, but will also be a DHCP in 10.1.10.1/24 range? I can then plug the Linksys into the GB switch, which is now picking up statics and the 10.1.10.1/24 ranges, tell the Linksys to use 10.1.10.5 or so. Now, do I disable DHCP on the Linksys, and the cable modem/firewall will pass through the statics and 10.0.10.1/24 ranges as well? Or, could I open a second DHCP pool on the Linksys? I guess doing so gives me network isolation again, but it is just the reverse of what I have now. But I get out of the bottleneck, not that the Linksys could ever really touch real GB speeds anyway, but the managed switch certainly can. This is all because 13 statics are not that many. Right now, 6 "servers", the Linksys, a managed switch, a few SSL certs, and I am running out. I do not want to waste a static IP on the managed GB switch, or the Linksys, unless it provides me some type of benefit. Final question, under my current setup, if I am on a workstation, sitting at 192.168.1.109, the Linksys, with GB, and I send a file over ssh to the static IP machine, is that literally leaving the internet, and coming back in, or does it stay local? To me it seems like: Workstation (192.168.1.109) -> Linksys DHCP -> Linksys Static IP -> Cable Modem -> Server ( and it hits the 10/100 ports on the cable modem, slowing me down. But does it round trip the network, leave and come back in, limiting me to the 50/10 internet speeds? *These are all made up numbers, I do not use default router IP's as I will one day add a VPN, and do not want collisions. I need some recommendations, do I want one big network, or two isolated ones. Printers these days need an IP, everything does, I can not get autoconf/bonjour to be reliable on most printers. but I am also not sure I want the "server" side of my operation to be polluted by the workstation side of my operation. Unless there is some magic subetting I have not learned yet, here is what I am thinking: Cable modem 10/100, has 13 static IP, publicly accessible -> Enable DHCP on the cable modem -> Cable modem plugs into managed switch -> Managed switch gets 10.1.10.1 ssh, telnet, https admin management address -> Managed switch sends static IP's to to servers -> Plug Linksys into managed switch, giving it 10.1.10.2 static internally in Linksys admin -> Linksys gets assigned 10.1.10.x as its DHCP sending range -> Local printers, workstations, iPhones etc, connect to this -> ( Do I enable DHCP or disable it on the Linksys, just define a non over lapping range, or create an entirely new DHCP at 10.1.50.0/24, I think I am back isolated again with that method too? ) Thank you for any suggestions. This is the first time I have had to deal with less than a /24, and most are larger than that, but it is just a drop to a cabinet. Otherwise, it's a router, a few repeaters, and soho stuff that is simple, with one IP. I know a few may suggest going all DHCP on the servers, and I may one day, just not now, there has been too much moving of gear for me to be interested in that, and I would want something in the Catalyst series to deal with that.

  Read the article

• How to diagnose failing 6Gbps SATA connection?

  - by whitequark

  I have a Samsung RC530 notebook and OCZ Vertex-3 6Gbps SATA SSD working in AHCI mode. # dmesg | grep DMI SAMSUNG ELECTRONICS CO., LTD. RC530/RC730/RC530/RC730, BIOS 03WD.M008.20110927.PSA 09/27/2011 # lspci -nn 00:1f.2 SATA controller [0106]: Intel Corporation 6 Series/C200 Series Chipset Family 6 port SATA AHCI Controller [8086:1c03] (rev 04) # sdparm -a /dev/sda /dev/sda: ATA OCZ-VERTEX3 2.15 At the boot, the following messages are present in dmesg (I am running Debian wheezy @ Linux 3.2.8): # dmesg | grep -iE '(ata|ahci)' [ 5.179783] ahci 0000:00:1f.2: version 3.0 [ 5.179802] ahci 0000:00:1f.2: PCI INT B -> GSI 19 (level, low) -> IRQ 19 [ 5.179864] ahci 0000:00:1f.2: irq 42 for MSI/MSI-X [ 5.195424] ahci 0000:00:1f.2: AHCI 0001.0300 32 slots 6 ports 6 Gbps 0x5 impl SATA mode [ 5.195429] ahci 0000:00:1f.2: flags: 64bit ncq sntf pm led clo pio slum part ems apst [ 5.195436] ahci 0000:00:1f.2: setting latency timer to 64 [ 5.204035] scsi0 : ahci [ 5.204301] scsi1 : ahci [ 5.204447] scsi2 : ahci [ 5.204592] scsi3 : ahci [ 5.204682] scsi4 : ahci [ 5.204799] scsi5 : ahci [ 5.204917] ata1: SATA max UDMA/133 abar m2048@0xf7c06000 port 0xf7c06100 irq 42 [ 5.204920] ata2: DUMMY [ 5.204923] ata3: SATA max UDMA/133 abar m2048@0xf7c06000 port 0xf7c06200 irq 42 [ 5.204924] ata4: DUMMY [ 5.204926] ata5: DUMMY [ 5.204927] ata6: DUMMY [ 5.523039] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300) [ 5.525911] ata3.00: ATAPI: TSSTcorp CDDVDW SN-208BB, SC00, max UDMA/100 [ 5.531006] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300) [ 5.533703] ata3.00: configured for UDMA/100 [ 5.542790] ata1.00: ATA-8: OCZ-VERTEX3, 2.15, max UDMA/133 [ 5.542800] ata1.00: 117231408 sectors, multi 16: LBA48 NCQ (depth 31/32), AA [ 5.552751] ata1.00: configured for UDMA/133 [ 5.553050] scsi 0:0:0:0: Direct-Access ATA OCZ-VERTEX3 2.15 PQ: 0 ANSI: 5 [ 5.559621] scsi 2:0:0:0: CD-ROM TSSTcorp CDDVDW SN-208BB SC00 PQ: 0 ANSI: 5 [ 5.564059] sd 0:0:0:0: [sda] 117231408 512-byte logical blocks: (60.0 GB/55.8 GiB) [ 5.564127] sd 0:0:0:0: [sda] Write Protect is off [ 5.564131] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00 [ 5.564158] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 5.564582] sda: sda1 [ 5.564810] sd 0:0:0:0: [sda] Attached SCSI disk [ 5.572006] sr0: scsi3-mmc drive: 16x/24x writer dvd-ram cd/rw xa/form2 cdda tray [ 5.572010] cdrom: Uniform CD-ROM driver Revision: 3.20 [ 5.572189] sr 2:0:0:0: Attached scsi CD-ROM sr0 [ 6.717181] ata1.00: exception Emask 0x50 SAct 0x1 SErr 0x280900 action 0x6 frozen [ 6.717238] ata1.00: irq_stat 0x08000000, interface fatal error [ 6.717291] ata1: SError: { UnrecovData HostInt 10B8B BadCRC } [ 6.717342] ata1.00: failed command: READ FPDMA QUEUED [ 6.717395] ata1.00: cmd 60/50:00:20:39:58/00:00:00:00:00/40 tag 0 ncq 40960 in [ 6.717396] res 40/00:00:20:39:58/00:00:00:00:00/40 Emask 0x50 (ATA bus error) [ 6.717503] ata1.00: status: { DRDY } [ 6.717553] ata1: hard resetting link [ 7.033417] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300) [ 7.055234] ata1.00: configured for UDMA/133 [ 7.055262] ata1: EH complete [ 7.147280] ata1.00: exception Emask 0x10 SAct 0xf8 SErr 0x280100 action 0x6 frozen [ 7.147340] ata1.00: irq_stat 0x08000000, interface fatal error [ 7.147393] ata1: SError: { UnrecovData 10B8B BadCRC } [ 7.147460] ata1.00: failed command: READ FPDMA QUEUED [ 7.147529] ata1.00: cmd 60/08:18:88:17:41/00:00:02:00:00/40 tag 3 ncq 4096 in [ 7.147531] res 40/00:38:50:99:64/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.147691] ata1.00: status: { DRDY } [ 7.147754] ata1.00: failed command: READ FPDMA QUEUED [ 7.147821] ata1.00: cmd 60/00:20:f8:42:4c/01:00:02:00:00/40 tag 4 ncq 131072 in [ 7.147822] res 40/00:38:50:99:64/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.147977] ata1.00: status: { DRDY } [ 7.148036] ata1.00: failed command: READ FPDMA QUEUED [ 7.148100] ata1.00: cmd 60/50:28:f8:43:4c/00:00:02:00:00/40 tag 5 ncq 40960 in [ 7.148101] res 40/00:38:50:99:64/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.148255] ata1.00: status: { DRDY } [ 7.148315] ata1.00: failed command: READ FPDMA QUEUED [ 7.148379] ata1.00: cmd 60/00:30:50:98:64/01:00:02:00:00/40 tag 6 ncq 131072 in [ 7.148380] res 40/00:38:50:99:64/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.148534] ata1.00: status: { DRDY } [ 7.148593] ata1.00: failed command: READ FPDMA QUEUED [ 7.148657] ata1.00: cmd 60/00:38:50:99:64/01:00:02:00:00/40 tag 7 ncq 131072 in [ 7.148658] res 40/00:38:50:99:64/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.148813] ata1.00: status: { DRDY } [ 7.148875] ata1: hard resetting link [ 7.464842] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300) [ 7.486794] ata1.00: configured for UDMA/133 [ 7.486822] ata1: EH complete [ 7.546395] ata1.00: exception Emask 0x10 SAct 0x2f SErr 0x280100 action 0x6 frozen [ 7.546470] ata1.00: irq_stat 0x08000000, interface fatal error [ 7.546531] ata1: SError: { UnrecovData 10B8B BadCRC } [ 7.546588] ata1.00: failed command: READ FPDMA QUEUED [ 7.546648] ata1.00: cmd 60/00:00:e0:4b:61/01:00:02:00:00/40 tag 0 ncq 131072 in [ 7.546649] res 40/00:28:e0:4c:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.546794] ata1.00: status: { DRDY } [ 7.546847] ata1.00: failed command: READ FPDMA QUEUED [ 7.546906] ata1.00: cmd 60/00:08:90:2f:48/01:00:02:00:00/40 tag 1 ncq 131072 in [ 7.546907] res 40/00:28:e0:4c:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.547053] ata1.00: status: { DRDY } [ 7.547106] ata1.00: failed command: READ FPDMA QUEUED [ 7.547165] ata1.00: cmd 60/00:10:90:30:48/01:00:02:00:00/40 tag 2 ncq 131072 in [ 7.547166] res 40/00:28:e0:4c:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.547310] ata1.00: status: { DRDY } [ 7.547363] ata1.00: failed command: READ FPDMA QUEUED [ 7.547422] ata1.00: cmd 60/00:18:50:c7:64/01:00:02:00:00/40 tag 3 ncq 131072 in [ 7.547423] res 40/00:28:e0:4c:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.547568] ata1.00: status: { DRDY } [ 7.547621] ata1.00: failed command: READ FPDMA QUEUED [ 7.547681] ata1.00: cmd 60/00:28:e0:4c:61/01:00:02:00:00/40 tag 5 ncq 131072 in [ 7.547682] res 40/00:28:e0:4c:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.547825] ata1.00: status: { DRDY } [ 7.547882] ata1: hard resetting link [ 7.864408] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300) [ 7.886351] ata1.00: configured for UDMA/133 [ 7.886375] ata1: EH complete [ 7.890012] ata1: limiting SATA link speed to 3.0 Gbps [ 7.890016] ata1.00: exception Emask 0x10 SAct 0x7 SErr 0x280100 action 0x6 frozen [ 7.890093] ata1.00: irq_stat 0x08000000, interface fatal error [ 7.890152] ata1: SError: { UnrecovData 10B8B BadCRC } [ 7.890210] ata1.00: failed command: READ FPDMA QUEUED [ 7.890272] ata1.00: cmd 60/00:00:90:33:48/01:00:02:00:00/40 tag 0 ncq 131072 in [ 7.890273] res 40/00:10:e0:4f:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.890418] ata1.00: status: { DRDY } [ 7.890472] ata1.00: failed command: READ FPDMA QUEUED [ 7.890530] ata1.00: cmd 60/00:08:90:34:48/01:00:02:00:00/40 tag 1 ncq 131072 in [ 7.890531] res 40/00:10:e0:4f:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.890672] ata1.00: status: { DRDY } [ 7.890724] ata1.00: failed command: READ FPDMA QUEUED [ 7.890781] ata1.00: cmd 60/78:10:e0:4f:61/00:00:02:00:00/40 tag 2 ncq 61440 in [ 7.890782] res 40/00:10:e0:4f:61/00:00:02:00:00/40 Emask 0x10 (ATA bus error) [ 7.890925] ata1.00: status: { DRDY } [ 7.890981] ata1: hard resetting link [ 8.208021] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 320) [ 8.230100] ata1.00: configured for UDMA/133 [ 8.230124] ata1: EH complete Looks like the SATA interface tries to use 6Gbps link, then fails miserably and Linux fallbacks to 3Gbps. This is somewhat fine for me, as the system boots successfully each time and works under high load (cd linux-3.2.8; make -j16). I've also ran memtest86+ and it did not find any errors. What concerns me more is that Grub sometimes takes a long time to load the images and/or fails to load itself completely. The error is consistent and is probablistic: that is, each time I boot I have a certain chance to fail. Actually, I have a slight suspiction on the cause of the failure. Look at the cabling: What kind of engineer does it this way? Nah. Even 1Gbps Ethernet hardly tolerates cables bent over a small angle, and there you have 6Gbps SATA. How cound I determine and fix the cause of errors and/or switch the link to 3Gbps mode permanently?

  Read the article

• Improving Partitioned Table Join Performance

  - by Paul White

  The query optimizer does not always choose an optimal strategy when joining partitioned tables. This post looks at an example, showing how a manual rewrite of the query can almost double performance, while reducing the memory grant to almost nothing. Test Data The two tables in this example use a common partitioning partition scheme. The partition function uses 41 equal-size partitions: CREATE PARTITION FUNCTION PFT (integer) AS RANGE RIGHT FOR VALUES ( 125000, 250000, 375000, 500000, 625000, 750000, 875000, 1000000, 1125000, 1250000, 1375000, 1500000, 1625000, 1750000, 1875000, 2000000, 2125000, 2250000, 2375000, 2500000, 2625000, 2750000, 2875000, 3000000, 3125000, 3250000, 3375000, 3500000, 3625000, 3750000, 3875000, 4000000, 4125000, 4250000, 4375000, 4500000, 4625000, 4750000, 4875000, 5000000 ); GO CREATE PARTITION SCHEME PST AS PARTITION PFT ALL TO ([PRIMARY]); There two tables are: CREATE TABLE dbo.T1 ( TID integer NOT NULL IDENTITY(0,1), Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T1 PRIMARY KEY CLUSTERED (TID) ON PST (TID) );   CREATE TABLE dbo.T2 ( TID integer NOT NULL, Column1 integer NOT NULL, Padding binary(100) NOT NULL DEFAULT 0x,   CONSTRAINT PK_T2 PRIMARY KEY CLUSTERED (TID, Column1) ON PST (TID) ); The next script loads 5 million rows into T1 with a pseudo-random value between 1 and 5 for Column1. The table is partitioned on the IDENTITY column TID: INSERT dbo.T1 WITH (TABLOCKX) (Column1) SELECT (ABS(CHECKSUM(NEWID())) % 5) + 1 FROM dbo.Numbers AS N WHERE n BETWEEN 1 AND 5000000; In case you don’t already have an auxiliary table of numbers lying around, here’s a script to create one with 10 million rows: CREATE TABLE dbo.Numbers (n bigint PRIMARY KEY);   WITH L0 AS(SELECT 1 AS c UNION ALL SELECT 1), L1 AS(SELECT 1 AS c FROM L0 AS A CROSS JOIN L0 AS B), L2 AS(SELECT 1 AS c FROM L1 AS A CROSS JOIN L1 AS B), L3 AS(SELECT 1 AS c FROM L2 AS A CROSS JOIN L2 AS B), L4 AS(SELECT 1 AS c FROM L3 AS A CROSS JOIN L3 AS B), L5 AS(SELECT 1 AS c FROM L4 AS A CROSS JOIN L4 AS B), Nums AS(SELECT ROW_NUMBER() OVER (ORDER BY (SELECT NULL)) AS n FROM L5) INSERT dbo.Numbers WITH (TABLOCKX) SELECT TOP (10000000) n FROM Nums ORDER BY n OPTION (MAXDOP 1); Table T1 contains data like this: Next we load data into table T2. The relationship between the two tables is that table 2 contains ‘n’ rows for each row in table 1, where ‘n’ is determined by the value in Column1 of table T1. There is nothing particularly special about the data or distribution, by the way. INSERT dbo.T2 WITH (TABLOCKX) (TID, Column1) SELECT T.TID, N.n FROM dbo.T1 AS T JOIN dbo.Numbers AS N ON N.n >= 1 AND N.n <= T.Column1; Table T2 ends up containing about 15 million rows: The primary key for table T2 is a combination of TID and Column1. The data is partitioned according to the value in column TID alone. Partition Distribution The following query shows the number of rows in each partition of table T1: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T1 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are 40 partitions containing 125,000 rows (40 * 125k = 5m rows). The rightmost partition remains empty. The next query shows the distribution for table 2: SELECT PartitionID = CA1.P, NumRows = COUNT_BIG(*) FROM dbo.T2 AS T CROSS APPLY (VALUES ($PARTITION.PFT(TID))) AS CA1 (P) GROUP BY CA1.P ORDER BY CA1.P; There are roughly 375,000 rows in each partition (the rightmost partition is also empty): Ok, that’s the test data done. Test Query and Execution Plan The task is to count the rows resulting from joining tables 1 and 2 on the TID column: SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; The optimizer chooses a plan using parallel hash join, and partial aggregation: The Plan Explorer plan tree view shows accurate cardinality estimates and an even distribution of rows across threads (click to enlarge the image): With a warm data cache, the STATISTICS IO output shows that no physical I/O was needed, and all 41 partitions were touched: Running the query without actual execution plan or STATISTICS IO information for maximum performance, the query returns in around 2600ms. Execution Plan Analysis The first step toward improving on the execution plan produced by the query optimizer is to understand how it works, at least in outline. The two parallel Clustered Index Scans use multiple threads to read rows from tables T1 and T2. Parallel scan uses a demand-based scheme where threads are given page(s) to scan from the table as needed. This arrangement has certain important advantages, but does result in an unpredictable distribution of rows amongst threads. The point is that multiple threads cooperate to scan the whole table, but it is impossible to predict which rows end up on which threads. For correct results from the parallel hash join, the execution plan has to ensure that rows from T1 and T2 that might join are processed on the same thread. For example, if a row from T1 with join key value ‘1234’ is placed in thread 5’s hash table, the execution plan must guarantee that any rows from T2 that also have join key value ‘1234’ probe thread 5’s hash table for matches. The way this guarantee is enforced in this parallel hash join plan is by repartitioning rows to threads after each parallel scan. The two repartitioning exchanges route rows to threads using a hash function over the hash join keys. The two repartitioning exchanges use the same hash function so rows from T1 and T2 with the same join key must end up on the same hash join thread. Expensive Exchanges This business of repartitioning rows between threads can be very expensive, especially if a large number of rows is involved. The execution plan selected by the optimizer moves 5 million rows through one repartitioning exchange and around 15 million across the other. As a first step toward removing these exchanges, consider the execution plan selected by the optimizer if we join just one partition from each table, disallowing parallelism: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = 1 AND $PARTITION.PFT(T2.TID) = 1 OPTION (MAXDOP 1); The optimizer has chosen a (one-to-many) merge join instead of a hash join. The single-partition query completes in around 100ms. If everything scaled linearly, we would expect that extending this strategy to all 40 populated partitions would result in an execution time around 4000ms. Using parallelism could reduce that further, perhaps to be competitive with the parallel hash join chosen by the optimizer. This raises a question. If the most efficient way to join one partition from each of the tables is to use a merge join, why does the optimizer not choose a merge join for the full query? Forcing a Merge Join Let’s force the optimizer to use a merge join on the test query using a hint: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN); This is the execution plan selected by the optimizer: This plan results in the same number of logical reads reported previously, but instead of 2600ms the query takes 5000ms. The natural explanation for this drop in performance is that the merge join plan is only using a single thread, whereas the parallel hash join plan could use multiple threads. Parallel Merge Join We can get a parallel merge join plan using the same query hint as before, and adding trace flag 8649: SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (MERGE JOIN, QUERYTRACEON 8649); The execution plan is: This looks promising. It uses a similar strategy to distribute work across threads as seen for the parallel hash join. In practice though, performance is disappointing. On a typical run, the parallel merge plan runs for around 8400ms; slower than the single-threaded merge join plan (5000ms) and much worse than the 2600ms for the parallel hash join. We seem to be going backwards! The logical reads for the parallel merge are still exactly the same as before, with no physical IOs. The cardinality estimates and thread distribution are also still very good (click to enlarge): A big clue to the reason for the poor performance is shown in the wait statistics (captured by Plan Explorer Pro): CXPACKET waits require careful interpretation, and are most often benign, but in this case excessive waiting occurs at the repartitioning exchanges. Unlike the parallel hash join, the repartitioning exchanges in this plan are order-preserving ‘merging’ exchanges (because merge join requires ordered inputs): Parallelism works best when threads can just grab any available unit of work and get on with processing it. Preserving order introduces inter-thread dependencies that can easily lead to significant waits occurring. In extreme cases, these dependencies can result in an intra-query deadlock, though the details of that will have to wait for another time to explore in detail. The potential for waits and deadlocks leads the query optimizer to cost parallel merge join relatively highly, especially as the degree of parallelism (DOP) increases. This high costing resulted in the optimizer choosing a serial merge join rather than parallel in this case. The test results certainly confirm its reasoning. Collocated Joins In SQL Server 2008 and later, the optimizer has another available strategy when joining tables that share a common partition scheme. This strategy is a collocated join, also known as as a per-partition join. It can be applied in both serial and parallel execution plans, though it is limited to 2-way joins in the current optimizer. Whether the optimizer chooses a collocated join or not depends on cost estimation. The primary benefits of a collocated join are that it eliminates an exchange and requires less memory, as we will see next. Costing and Plan Selection The query optimizer did consider a collocated join for our original query, but it was rejected on cost grounds. The parallel hash join with repartitioning exchanges appeared to be a cheaper option. There is no query hint to force a collocated join, so we have to mess with the costing framework to produce one for our test query. Pretending that IOs cost 50 times more than usual is enough to convince the optimizer to use collocated join with our test query: -- Pretend IOs are 50x cost temporarily DBCC SETIOWEIGHT(50);   -- Co-located hash join SELECT COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID OPTION (RECOMPILE);   -- Reset IO costing DBCC SETIOWEIGHT(1); Collocated Join Plan The estimated execution plan for the collocated join is: The Constant Scan contains one row for each partition of the shared partitioning scheme, from 1 to 41. The hash repartitioning exchanges seen previously are replaced by a single Distribute Streams exchange using Demand partitioning. Demand partitioning means that the next partition id is given to the next parallel thread that asks for one. My test machine has eight logical processors, and all are available for SQL Server to use. As a result, there are eight threads in the single parallel branch in this plan, each processing one partition from each table at a time. Once a thread finishes processing a partition, it grabs a new partition number from the Distribute Streams exchange…and so on until all partitions have been processed. It is important to understand that the parallel scans in this plan are different from the parallel hash join plan. Although the scans have the same parallelism icon, tables T1 and T2 are not being co-operatively scanned by multiple threads in the same way. Each thread reads a single partition of T1 and performs a hash match join with the same partition from table T2. The properties of the two Clustered Index Scans show a Seek Predicate (unusual for a scan!) limiting the rows to a single partition: The crucial point is that the join between T1 and T2 is on TID, and TID is the partitioning column for both tables. A thread that processes partition ‘n’ is guaranteed to see all rows that can possibly join on TID for that partition. In addition, no other thread will see rows from that partition, so this removes the need for repartitioning exchanges. CPU and Memory Efficiency Improvements The collocated join has removed two expensive repartitioning exchanges and added a single exchange processing 41 rows (one for each partition id). Remember, the parallel hash join plan exchanges had to process 5 million and 15 million rows. The amount of processor time spent on exchanges will be much lower in the collocated join plan. In addition, the collocated join plan has a maximum of 8 threads processing single partitions at any one time. The 41 partitions will all be processed eventually, but a new partition is not started until a thread asks for it. Threads can reuse hash table memory for the new partition. The parallel hash join plan also had 8 hash tables, but with all 5,000,000 build rows loaded at the same time. The collocated plan needs memory for only 8 * 125,000 = 1,000,000 rows at any one time. Collocated Hash Join Performance The collated join plan has disappointing performance in this case. The query runs for around 25,300ms despite the same IO statistics as usual. This is much the worst result so far, so what went wrong? It turns out that cardinality estimation for the single partition scans of table T1 is slightly low. The properties of the Clustered Index Scan of T1 (graphic immediately above) show the estimation was for 121,951 rows. This is a small shortfall compared with the 125,000 rows actually encountered, but it was enough to cause the hash join to spill to physical tempdb: A level 1 spill doesn’t sound too bad, until you realize that the spill to tempdb probably occurs for each of the 41 partitions. As a side note, the cardinality estimation error is a little surprising because the system tables accurately show there are 125,000 rows in every partition of T1. Unfortunately, the optimizer uses regular column and index statistics to derive cardinality estimates here rather than system table information (e.g. sys.partitions). Collocated Merge Join We will never know how well the collocated parallel hash join plan might have worked without the cardinality estimation error (and the resulting 41 spills to tempdb) but we do know: Merge join does not require a memory grant; and Merge join was the optimizer’s preferred join option for a single partition join Putting this all together, what we would really like to see is the same collocated join strategy, but using merge join instead of hash join. Unfortunately, the current query optimizer cannot produce a collocated merge join; it only knows how to do collocated hash join. So where does this leave us? CROSS APPLY sys.partitions We can try to write our own collocated join query. We can use sys.partitions to find the partition numbers, and CROSS APPLY to get a count per partition, with a final step to sum the partial counts. The following query implements this idea: SELECT row_count = SUM(Subtotals.cnt) FROM ( -- Partition numbers SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1 ) AS P CROSS APPLY ( -- Count per collocated join SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals; The estimated plan is: The cardinality estimates aren’t all that good here, especially the estimate for the scan of the system table underlying the sys.partitions view. Nevertheless, the plan shape is heading toward where we would like to be. Each partition number from the system table results in a per-partition scan of T1 and T2, a one-to-many Merge Join, and a Stream Aggregate to compute the partial counts. The final Stream Aggregate just sums the partial counts. Execution time for this query is around 3,500ms, with the same IO statistics as always. This compares favourably with 5,000ms for the serial plan produced by the optimizer with the OPTION (MERGE JOIN) hint. This is another case of the sum of the parts being less than the whole – summing 41 partial counts from 41 single-partition merge joins is faster than a single merge join and count over all partitions. Even so, this single-threaded collocated merge join is not as quick as the original parallel hash join plan, which executed in 2,600ms. On the positive side, our collocated merge join uses only one logical processor and requires no memory grant. The parallel hash join plan used 16 threads and reserved 569 MB of memory:   Using a Temporary Table Our collocated merge join plan should benefit from parallelism. The reason parallelism is not being used is that the query references a system table. We can work around that by writing the partition numbers to a temporary table (or table variable): SET STATISTICS IO ON; DECLARE @s datetime2 = SYSUTCDATETIME();   CREATE TABLE #P ( partition_number integer PRIMARY KEY);   INSERT #P (partition_number) SELECT p.partition_number FROM sys.partitions AS p WHERE p.[object_id] = OBJECT_ID(N'T1', N'U') AND p.index_id = 1;   SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals;   DROP TABLE #P;   SELECT DATEDIFF(Millisecond, @s, SYSUTCDATETIME()); SET STATISTICS IO OFF; Using the temporary table adds a few logical reads, but the overall execution time is still around 3500ms, indistinguishable from the same query without the temporary table. The problem is that the query optimizer still doesn’t choose a parallel plan for this query, though the removal of the system table reference means that it could if it chose to: In fact the optimizer did enter the parallel plan phase of query optimization (running search 1 for a second time): Unfortunately, the parallel plan found seemed to be more expensive than the serial plan. This is a crazy result, caused by the optimizer’s cost model not reducing operator CPU costs on the inner side of a nested loops join. Don’t get me started on that, we’ll be here all night. In this plan, everything expensive happens on the inner side of a nested loops join. Without a CPU cost reduction to compensate for the added cost of exchange operators, candidate parallel plans always look more expensive to the optimizer than the equivalent serial plan. Parallel Collocated Merge Join We can produce the desired parallel plan using trace flag 8649 again: SELECT row_count = SUM(Subtotals.cnt) FROM #P AS p CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: One difference between this plan and the collocated hash join plan is that a Repartition Streams exchange operator is used instead of Distribute Streams. The effect is similar, though not quite identical. The Repartition uses round-robin partitioning, meaning the next partition id is pushed to the next thread in sequence. The Distribute Streams exchange seen earlier used Demand partitioning, meaning the next partition id is pulled across the exchange by the next thread that is ready for more work. There are subtle performance implications for each partitioning option, but going into that would again take us too far off the main point of this post. Performance The important thing is the performance of this parallel collocated merge join – just 1350ms on a typical run. The list below shows all the alternatives from this post (all timings include creation, population, and deletion of the temporary table where appropriate) from quickest to slowest: Collocated parallel merge join: 1350ms Parallel hash join: 2600ms Collocated serial merge join: 3500ms Serial merge join: 5000ms Parallel merge join: 8400ms Collated parallel hash join: 25,300ms (hash spill per partition) The parallel collocated merge join requires no memory grant (aside from a paltry 1.2MB used for exchange buffers). This plan uses 16 threads at DOP 8; but 8 of those are (rather pointlessly) allocated to the parallel scan of the temporary table. These are minor concerns, but it turns out there is a way to address them if it bothers you. Parallel Collocated Merge Join with Demand Partitioning This final tweak replaces the temporary table with a hard-coded list of partition ids (dynamic SQL could be used to generate this query from sys.partitions): SELECT row_count = SUM(Subtotals.cnt) FROM ( VALUES (1),(2),(3),(4),(5),(6),(7),(8),(9),(10), (11),(12),(13),(14),(15),(16),(17),(18),(19),(20), (21),(22),(23),(24),(25),(26),(27),(28),(29),(30), (31),(32),(33),(34),(35),(36),(37),(38),(39),(40),(41) ) AS P (partition_number) CROSS APPLY ( SELECT cnt = COUNT_BIG(*) FROM dbo.T1 AS T1 JOIN dbo.T2 AS T2 ON T2.TID = T1.TID WHERE $PARTITION.PFT(T1.TID) = p.partition_number AND $PARTITION.PFT(T2.TID) = p.partition_number ) AS SubTotals OPTION (QUERYTRACEON 8649); The actual execution plan is: The parallel collocated hash join plan is reproduced below for comparison: The manual rewrite has another advantage that has not been mentioned so far: the partial counts (per partition) can be computed earlier than the partial counts (per thread) in the optimizer’s collocated join plan. The earlier aggregation is performed by the extra Stream Aggregate under the nested loops join. The performance of the parallel collocated merge join is unchanged at around 1350ms. Final Words It is a shame that the current query optimizer does not consider a collocated merge join (Connect item closed as Won’t Fix). The example used in this post showed an improvement in execution time from 2600ms to 1350ms using a modestly-sized data set and limited parallelism. In addition, the memory requirement for the query was almost completely eliminated  – down from 569MB to 1.2MB. The problem with the parallel hash join selected by the optimizer is that it attempts to process the full data set all at once (albeit using eight threads). It requires a large memory grant to hold all 5 million rows from table T1 across the eight hash tables, and does not take advantage of the divide-and-conquer opportunity offered by the common partitioning. The great thing about the collocated join strategies is that each parallel thread works on a single partition from both tables, reading rows, performing the join, and computing a per-partition subtotal, before moving on to a new partition. From a thread’s point of view… If you have trouble visualizing what is happening from just looking at the parallel collocated merge join execution plan, let’s look at it again, but from the point of view of just one thread operating between the two Parallelism (exchange) operators. Our thread picks up a single partition id from the Distribute Streams exchange, and starts a merge join using ordered rows from partition 1 of table T1 and partition 1 of table T2. By definition, this is all happening on a single thread. As rows join, they are added to a (per-partition) count in the Stream Aggregate immediately above the Merge Join. Eventually, either T1 (partition 1) or T2 (partition 1) runs out of rows and the merge join stops. The per-partition count from the aggregate passes on through the Nested Loops join to another Stream Aggregate, which is maintaining a per-thread subtotal. Our same thread now picks up a new partition id from the exchange (say it gets id 9 this time). The count in the per-partition aggregate is reset to zero, and the processing of partition 9 of both tables proceeds just as it did for partition 1, and on the same thread. Each thread picks up a single partition id and processes all the data for that partition, completely independently from other threads working on other partitions. One thread might eventually process partitions (1, 9, 17, 25, 33, 41) while another is concurrently processing partitions (2, 10, 18, 26, 34) and so on for the other six threads at DOP 8. The point is that all 8 threads can execute independently and concurrently, continuing to process new partitions until the wider job (of which the thread has no knowledge!) is done. This divide-and-conquer technique can be much more efficient than simply splitting the entire workload across eight threads all at once. Related Reading Understanding and Using Parallelism in SQL Server Parallel Execution Plans Suck © 2013 Paul White – All Rights Reserved Twitter: @SQL_Kiwi

  Read the article

• Scrum in 5 Minutes

  - by Stephen.Walther

  The goal of this blog entry is to explain the basic concepts of Scrum in less than five minutes. You learn how Scrum can help a team of developers to successfully complete a complex software project. Product Backlog and the Product Owner Imagine that you are part of a team which needs to create a new website – for example, an e-commerce website. You have an overwhelming amount of work to do. You need to build (or possibly buy) a shopping cart, install an SSL certificate, create a product catalog, create a Facebook page, and at least a hundred other things that you have not thought of yet. According to Scrum, the first thing you should do is create a list. Place the highest priority items at the top of the list and the lower priority items lower in the list. For example, creating the shopping cart and buying the domain name might be high priority items and creating a Facebook page might be a lower priority item. In Scrum, this list is called the Product Backlog. How do you prioritize the items in the Product Backlog? Different stakeholders in the project might have different priorities. Gary, your division VP, thinks that it is crucial that the e-commerce site has a mobile app. Sally, your direct manager, thinks taking advantage of new HTML5 features is much more important. Multiple people are pulling you in different directions. According to Scrum, it is important that you always designate one person, and only one person, as the Product Owner. The Product Owner is the person who decides what items should be added to the Product Backlog and the priority of the items in the Product Backlog. The Product Owner could be the customer who is paying the bills, the project manager who is responsible for delivering the project, or a customer representative. The critical point is that the Product Owner must always be a single person and that single person has absolute authority over the Product Backlog. Sprints and the Sprint Backlog So now the developer team has a prioritized list of items and they can start work. The team starts implementing the first item in the Backlog — the shopping cart — and the team is making good progress. Unfortunately, however, half-way through the work of implementing the shopping cart, the Product Owner changes his mind. The Product Owner decides that it is much more important to create the product catalog before the shopping cart. With some frustration, the team switches their developmental efforts to focus on implementing the product catalog. However, part way through completing this work, once again the Product Owner changes his mind about the highest priority item. Getting work done when priorities are constantly shifting is frustrating for the developer team and it results in lower productivity. At the same time, however, the Product Owner needs to have absolute authority over the priority of the items which need to get done. Scrum solves this conflict with the concept of Sprints. In Scrum, a developer team works in Sprints. At the beginning of a Sprint the developers and the Product Owner agree on the items from the backlog which they will complete during the Sprint. This subset of items from the Product Backlog becomes the Sprint Backlog. During the Sprint, the Product Owner is not allowed to change the items in the Sprint Backlog. In other words, the Product Owner cannot shift priorities on the developer team during the Sprint. Different teams use Sprints of different lengths such as one month Sprints, two-week Sprints, and one week Sprints. For high-stress, time critical projects, teams typically choose shorter sprints such as one week sprints. For more mature projects, longer one month sprints might be more appropriate. A team can pick whatever Sprint length makes sense for them just as long as the team is consistent. You should pick a Sprint length and stick with it. Daily Scrum During a Sprint, the developer team needs to have meetings to coordinate their work on completing the items in the Sprint Backlog. For example, the team needs to discuss who is working on what and whether any blocking issues have been discovered. Developers hate meetings (well, sane developers hate meetings). Meetings take developers away from their work of actually implementing stuff as opposed to talking about implementing stuff. However, a developer team which never has meetings and never coordinates their work also has problems. For example, Fred might get stuck on a programming problem for days and never reach out for help even though Tom (who sits in the cubicle next to him) has already solved the very same problem. Or, both Ted and Fred might have started working on the same item from the Sprint Backlog at the same time. In Scrum, these conflicting needs – limiting meetings but enabling team coordination – are resolved with the idea of the Daily Scrum. The Daily Scrum is a meeting for coordinating the work of the developer team which happens once a day. To keep the meeting short, each developer answers only the following three questions: 1. What have you done since yesterday? 2. What do you plan to do today? 3. Any impediments in your way? During the Daily Scrum, developers are not allowed to talk about issues with their cat, do demos of their latest work, or tell heroic stories of programming problems overcome. The meeting must be kept short — typically about 15 minutes. Issues which come up during the Daily Scrum should be discussed in separate meetings which do not involve the whole developer team. Stories and Tasks Items in the Product or Sprint Backlog – such as building a shopping cart or creating a Facebook page – are often referred to as User Stories or Stories. The Stories are created by the Product Owner and should represent some business need. Unlike the Product Owner, the developer team needs to think about how a Story should be implemented. At the beginning of a Sprint, the developer team takes the Stories from the Sprint Backlog and breaks the stories into tasks. For example, the developer team might take the Create a Shopping Cart story and break it into the following tasks: · Enable users to add and remote items from shopping cart · Persist the shopping cart to database between visits · Redirect user to checkout page when Checkout button is clicked During the Daily Scrum, members of the developer team volunteer to complete the tasks required to implement the next Story in the Sprint Backlog. When a developer talks about what he did yesterday or plans to do tomorrow then the developer should be referring to a task. Stories are owned by the Product Owner and a story is all about business value. In contrast, the tasks are owned by the developer team and a task is all about implementation details. A story might take several days or weeks to complete. A task is something which a developer can complete in less than a day. Some teams get lazy about breaking stories into tasks. Neglecting to break stories into tasks can lead to “Never Ending Stories” If you don’t break a story into tasks, then you can’t know how much of a story has actually been completed because you don’t have a clear idea about the implementation steps required to complete the story. Scrumboard During the Daily Scrum, the developer team uses a Scrumboard to coordinate their work. A Scrumboard contains a list of the stories for the current Sprint, the tasks associated with each Story, and the state of each task. The developer team uses the Scrumboard so everyone on the team can see, at a glance, what everyone is working on. As a developer works on a task, the task moves from state to state and the state of the task is updated on the Scrumboard. Common task states are ToDo, In Progress, and Done. Some teams include additional task states such as Needs Review or Needs Testing. Some teams use a physical Scrumboard. In that case, you use index cards to represent the stories and the tasks and you tack the index cards onto a physical board. Using a physical Scrumboard has several disadvantages. A physical Scrumboard does not work well with a distributed team – for example, it is hard to share the same physical Scrumboard between Boston and Seattle. Also, generating reports from a physical Scrumboard is more difficult than generating reports from an online Scrumboard. Estimating Stories and Tasks Stakeholders in a project, the people investing in a project, need to have an idea of how a project is progressing and when the project will be completed. For example, if you are investing in creating an e-commerce site, you need to know when the site can be launched. It is not enough to just say that “the project will be done when it is done” because the stakeholders almost certainly have a limited budget to devote to the project. The people investing in the project cannot determine the business value of the project unless they can have an estimate of how long it will take to complete the project. Developers hate to give estimates. The reason that developers hate to give estimates is that the estimates are almost always completely made up. For example, you really don’t know how long it takes to build a shopping cart until you finish building a shopping cart, and at that point, the estimate is no longer useful. The problem is that writing code is much more like Finding a Cure for Cancer than Building a Brick Wall. Building a brick wall is very straightforward. After you learn how to add one brick to a wall, you understand everything that is involved in adding a brick to a wall. There is no additional research required and no surprises. If, on the other hand, I assembled a team of scientists and asked them to find a cure for cancer, and estimate exactly how long it will take, they would have no idea. The problem is that there are too many unknowns. I don’t know how to cure cancer, I need to do a lot of research here, so I cannot even begin to estimate how long it will take. So developers hate to provide estimates, but the Product Owner and other product stakeholders, have a legitimate need for estimates. Scrum resolves this conflict by using the idea of Story Points. Different teams use different units to represent Story Points. For example, some teams use shirt sizes such as Small, Medium, Large, and X-Large. Some teams prefer to use Coffee Cup sizes such as Tall, Short, and Grande. Finally, some teams like to use numbers from the Fibonacci series. These alternative units are converted into a Story Point value. Regardless of the type of unit which you use to represent Story Points, the goal is the same. Instead of attempting to estimate a Story in hours (which is doomed to failure), you use a much less fine-grained measure of work. A developer team is much more likely to be able to estimate that a Story is Small or X-Large than the exact number of hours required to complete the story. So you can think of Story Points as a compromise between the needs of the Product Owner and the developer team. When a Sprint starts, the developer team devotes more time to thinking about the Stories in a Sprint and the developer team breaks the Stories into Tasks. In Scrum, you estimate the work required to complete a Story by using Story Points and you estimate the work required to complete a task by using hours. The difference between Stories and Tasks is that you don’t create a task until you are just about ready to start working on a task. A task is something that you should be able to create within a day, so you have a much better chance of providing an accurate estimate of the work required to complete a task than a story. Burndown Charts In Scrum, you use Burndown charts to represent the remaining work on a project. You use Release Burndown charts to represent the overall remaining work for a project and you use Sprint Burndown charts to represent the overall remaining work for a particular Sprint. You create a Release Burndown chart by calculating the remaining number of uncompleted Story Points for the entire Product Backlog every day. The vertical axis represents Story Points and the horizontal axis represents time. A Sprint Burndown chart is similar to a Release Burndown chart, but it focuses on the remaining work for a particular Sprint. There are two different types of Sprint Burndown charts. You can either represent the remaining work in a Sprint with Story Points or with task hours (the following image, taken from Wikipedia, uses hours). When each Product Backlog Story is completed, the Release Burndown chart slopes down. When each Story or task is completed, the Sprint Burndown chart slopes down. Burndown charts typically do not always slope down over time. As new work is added to the Product Backlog, the Release Burndown chart slopes up. If new tasks are discovered during a Sprint, the Sprint Burndown chart will also slope up. The purpose of a Burndown chart is to give you a way to track team progress over time. If, halfway through a Sprint, the Sprint Burndown chart is still climbing a hill then you know that you are in trouble. Team Velocity Stakeholders in a project always want more work done faster. For example, the Product Owner for the e-commerce site wants the website to launch before tomorrow. Developers tend to be overly optimistic. Rarely do developers acknowledge the physical limitations of reality. So Project stakeholders and the developer team often collude to delude themselves about how much work can be done and how quickly. Too many software projects begin in a state of optimism and end in frustration as deadlines zoom by. In Scrum, this problem is overcome by calculating a number called the Team Velocity. The Team Velocity is a measure of the average number of Story Points which a team has completed in previous Sprints. Knowing the Team Velocity is important during the Sprint Planning meeting when the Product Owner and the developer team work together to determine the number of stories which can be completed in the next Sprint. If you know the Team Velocity then you can avoid committing to do more work than the team has been able to accomplish in the past, and your team is much more likely to complete all of the work required for the next Sprint. Scrum Master There are three roles in Scrum: the Product Owner, the developer team, and the Scrum Master. I’v e already discussed the Product Owner. The Product Owner is the one and only person who maintains the Product Backlog and prioritizes the stories. I’ve also described the role of the developer team. The members of the developer team do the work of implementing the stories by breaking the stories into tasks. The final role, which I have not discussed, is the role of the Scrum Master. The Scrum Master is responsible for ensuring that the team is following the Scrum process. For example, the Scrum Master is responsible for making sure that there is a Daily Scrum meeting and that everyone answers the standard three questions. The Scrum Master is also responsible for removing (non-technical) impediments which the team might encounter. For example, if the team cannot start work until everyone installs the latest version of Microsoft Visual Studio then the Scrum Master has the responsibility of working with management to get the latest version of Visual Studio as quickly as possible. The Scrum Master can be a member of the developer team. Furthermore, different people can take on the role of the Scrum Master over time. The Scrum Master, however, cannot be the same person as the Product Owner. Using SonicAgile SonicAgile (SonicAgile.com) is an online tool which you can use to manage your projects using Scrum. You can use the SonicAgile Product Backlog to create a prioritized list of stories. You can estimate the size of the Stories using different Story Point units such as Shirt Sizes and Coffee Cup sizes. You can use SonicAgile during the Sprint Planning meeting to select the Stories that you want to complete during a particular Sprint. You can configure Sprints to be any length of time. SonicAgile calculates Team Velocity automatically and displays a warning when you add too many stories to a Sprint. In other words, it warns you when it thinks you are overcommitting in a Sprint. SonicAgile also includes a Scrumboard which displays the list of Stories selected for a Sprint and the tasks associated with each story. You can drag tasks from one task state to another. Finally, SonicAgile enables you to generate Release Burndown and Sprint Burndown charts. You can use these charts to view the progress of your team. To learn more about SonicAgile, visit SonicAgile.com. Summary In this post, I described many of the basic concepts of Scrum. You learned how a Product Owner uses a Product Backlog to create a prioritized list of tasks. I explained why work is completed in Sprints so the developer team can be more productive. I also explained how a developer team uses the daily scrum to coordinate their work. You learned how the developer team uses a Scrumboard to see, at a glance, who is working on what and the state of each task. I also discussed Burndown charts. You learned how you can use both Release and Sprint Burndown charts to track team progress in completing a project. Finally, I described the crucial role of the Scrum Master – the person who is responsible for ensuring that the rules of Scrum are being followed. My goal was not to describe all of the concepts of Scrum. This post was intended to be an introductory overview. For a comprehensive explanation of Scrum, I recommend reading Ken Schwaber’s book Agile Project Management with Scrum: http://www.amazon.com/Agile-Project-Management-Microsoft-Professional/dp/073561993X/ref=la_B001H6ODMC_1_1?ie=UTF8&qid=1345224000&sr=1-1

  Read the article

• Quick guide to Oracle IRM 11g: Classification design

  - by Simon Thorpe

  Quick guide to Oracle IRM 11g indexThis is the final article in the quick guide to Oracle IRM. If you've followed everything prior you will now have a fully functional and tested Information Rights Management service. It doesn't matter if you've been following the 10g or 11g guide as this next article is common to both. ContentsWhy this is the most important part... Understanding the classification and standard rights model Identifying business use cases Creating an effective IRM classification modelOne single classification across the entire businessA context for each and every possible granular use caseWhat makes a good context? Deciding on the use of roles in the context Reviewing the features and security for context roles Summary Why this is the most important part...Now the real work begins, installing and getting an IRM system running is as simple as following instructions. However to actually have an IRM technology easily protecting your most sensitive information without interfering with your users existing daily work flows and be able to scale IRM across the entire business, requires thought into how confidential documents are created, used and distributed. This article is going to give you the information you need to ask the business the right questions so that you can deploy your IRM service successfully. The IRM team here at Oracle have over 10 years of experience in helping customers and it is important you understand the following to be successful in securing access to your most confidential information. Whatever you are trying to secure, be it mergers and acquisitions information, engineering intellectual property, health care documentation or financial reports. No matter what type of user is going to access the information, be they employees, contractors or customers, there are common goals you are always trying to achieve.Securing the content at the earliest point possible and do it automatically. Removing the dependency on the user to decide to secure the content reduces the risk of mistakes significantly and therefore results a more secure deployment. K.I.S.S. (Keep It Simple Stupid) Reduce complexity in the rights/classification model. Oracle IRM lets you make changes to access to documents even after they are secured which allows you to start with a simple model and then introduce complexity once you've understood how the technology is going to be used in the business. After an initial learning period you can review your implementation and start to make informed decisions based on user feedback and administration experience. Clearly communicate to the user, when appropriate, any changes to their existing work practice. You must make every effort to make the transition to sealed content as simple as possible. For external users you must help them understand why you are securing the documents and inform them the value of the technology to both your business and them. Before getting into the detail, I must pay homage to Martin White, Vice President of client services in SealedMedia, the company Oracle acquired and who created Oracle IRM. In the SealedMedia years Martin was involved with every single customer and was key to the design of certain aspects of the IRM technology, specifically the context model we will be discussing here. Listening carefully to customers and understanding the flexibility of the IRM technology, Martin taught me all the skills of helping customers build scalable, effective and simple to use IRM deployments. No matter how well the engineering department designed the software, badly designed and poorly executed projects can result in difficult to use and manage, and ultimately insecure solutions. The advice and information that follows was born with Martin and he's still delivering IRM consulting with customers and can be found at www.thinkers.co.uk. It is from Martin and others that Oracle not only has the most advanced, scalable and usable document security solution on the market, but Oracle and their partners have the most experience in delivering successful document security solutions. Understanding the classification and standard rights model The goal of any successful IRM deployment is to balance the increase in security the technology brings without over complicating the way people use secured content and avoid a significant increase in administration and maintenance. With Oracle it is possible to automate the protection of content, deploy the desktop software transparently and use authentication methods such that users can open newly secured content initially unaware the document is any different to an insecure one. That is until of course they attempt to do something for which they don't have any rights, such as copy and paste to an insecure application or try and print. Central to achieving this objective is creating a classification model that is simple to understand and use but also provides the right level of complexity to meet the business needs. In Oracle IRM the term used for each classification is a "context". A context defines the relationship between.A group of related documents The people that use the documents The roles that these people perform The rights that these people need to perform their role The context is the key to the success of Oracle IRM. It provides the separation of the role and rights of a user from the content itself. Documents are sealed to contexts but none of the rights, user or group information is stored within the content itself. Sealing only places information about the location of the IRM server that sealed it, the context applied to the document and a few other pieces of metadata that pertain only to the document. This important separation of rights from content means that millions of documents can be secured against a single classification and a user needs only one right assigned to be able to access all documents. If you have followed all the previous articles in this guide, you will be ready to start defining contexts to which your sensitive information will be protected. But before you even start with IRM, you need to understand how your own business uses and creates sensitive documents and emails. Identifying business use cases Oracle is able to support multiple classification systems, but usually there is one single initial need for the technology which drives a deployment. This need might be to protect sensitive mergers and acquisitions information, engineering intellectual property, financial documents. For this and every subsequent use case you must understand how users create and work with documents, to who they are distributed and how the recipients should interact with them. A successful IRM deployment should start with one well identified use case (we go through some examples towards the end of this article) and then after letting this use case play out in the business, you learn how your users work with content, how well your communication to the business worked and if the classification system you deployed delivered the right balance. It is at this point you can start rolling the technology out further. Creating an effective IRM classification model Once you have selected the initial use case you will address with IRM, you need to design a classification model that defines the access to secured documents within the use case. In Oracle IRM there is an inbuilt classification system called the "context" model. In Oracle IRM 11g it is possible to extend the server to support any rights classification model, but the majority of users who are not using an application integration (such as Oracle IRM within Oracle Beehive) are likely to be starting out with the built in context model. Before looking at creating a classification system with IRM, it is worth reviewing some recognized standards and methods for creating and implementing security policy. A very useful set of documents are the ISO 17799 guidelines and the SANS security policy templates. First task is to create a context against which documents are to be secured. A context consists of a group of related documents (all top secret engineering research), a list of roles (contributors and readers) which define how users can access documents and a list of users (research engineers) who have been given a role allowing them to interact with sealed content. Before even creating the first context it is wise to decide on a philosophy which will dictate the level of granularity, the question is, where do you start? At a department level? By project? By technology? First consider the two ends of the spectrum... One single classification across the entire business Imagine that instead of having separate contexts, one for engineering intellectual property, one for your financial data, one for human resources personally identifiable information, you create one context for all documents across the entire business. Whilst you may have immediate objections, there are some significant benefits in thinking about considering this. Document security classification decisions are simple. You only have one context to chose from! User provisioning is simple, just make sure everyone has a role in the only context in the business. Administration is very low, if you assign rights to groups from the business user repository you probably never have to touch IRM administration again. There are however some obvious downsides to this model.All users in have access to all IRM secured content. So potentially a sales person could access sensitive mergers and acquisition documents, if they can get their hands on a copy that is. You cannot delegate control of different documents to different parts of the business, this may not satisfy your regulatory requirements for the separation and delegation of duties. Changing a users role affects every single document ever secured. Even though it is very unlikely a business would ever use one single context to secure all their sensitive information, thinking about this scenario raises one very important point. Just having one single context and securing all confidential documents to it, whilst incurring some of the problems detailed above, has one huge value. Once secured, IRM protected content can ONLY be accessed by authorized users. Just think of all the sensitive documents in your business today, imagine if you could ensure that only everyone you trust could open them. Even if an employee lost a laptop or someone accidentally sent an email to the wrong recipient, only the right people could open that file. A context for each and every possible granular use case Now let's think about the total opposite of a single context design. What if you created a context for each and every single defined business need and created multiple contexts within this for each level of granularity? Let's take a use case where we need to protect engineering intellectual property. Imagine we have 6 different engineering groups, and in each we have a research department, a design department and manufacturing. The company information security policy defines 3 levels of information sensitivity... restricted, confidential and top secret. Then let's say that each group and department needs to define access to information from both internal and external users. Finally add into the mix that they want to review the rights model for each context every financial quarter. This would result in a huge amount of contexts. For example, lets just look at the resulting contexts for one engineering group. Q1FY2010 Restricted Internal - Engineering Group 1 - Research Q1FY2010 Restricted Internal - Engineering Group 1 - Design Q1FY2010 Restricted Internal - Engineering Group 1 - Manufacturing Q1FY2010 Restricted External- Engineering Group 1 - Research Q1FY2010 Restricted External - Engineering Group 1 - Design Q1FY2010 Restricted External - Engineering Group 1 - Manufacturing Q1FY2010 Confidential Internal - Engineering Group 1 - Research Q1FY2010 Confidential Internal - Engineering Group 1 - Design Q1FY2010 Confidential Internal - Engineering Group 1 - Manufacturing Q1FY2010 Confidential External - Engineering Group 1 - Research Q1FY2010 Confidential External - Engineering Group 1 - Design Q1FY2010 Confidential External - Engineering Group 1 - Manufacturing Q1FY2010 Top Secret Internal - Engineering Group 1 - Research Q1FY2010 Top Secret Internal - Engineering Group 1 - Design Q1FY2010 Top Secret Internal - Engineering Group 1 - Manufacturing Q1FY2010 Top Secret External - Engineering Group 1 - Research Q1FY2010 Top Secret External - Engineering Group 1 - Design Q1FY2010 Top Secret External - Engineering Group 1 - Manufacturing Now multiply the above by 6 for each engineering group, 18 contexts. You are then creating/reviewing another 18 every 3 months. After a year you've got 72 contexts. What would be the advantages of such a complex classification model? You can satisfy very granular rights requirements, for example only an authorized engineering group 1 researcher can create a top secret report for access internally, and his role will be reviewed on a very frequent basis. Your business may have very complex rights requirements and mapping this directly to IRM may be an obvious exercise. The disadvantages of such a classification model are significant...Huge administrative overhead. Someone in the business must manage, review and administrate each of these contexts. If the engineering group had a single administrator, they would have 72 classifications to reside over each year. From an end users perspective life will be very confusing. Imagine if a user has rights in just 6 of these contexts. They may be able to print content from one but not another, be able to edit content in 2 contexts but not the other 4. Such confusion at the end user level causes frustration and resistance to the use of the technology. Increased synchronization complexity. Imagine a user who after 3 years in the company ends up with over 300 rights in many different contexts across the business. This would result in long synchronization times as the client software updates all your offline rights. Hard to understand who can do what with what. Imagine being the VP of engineering and as part of an internal security audit you are asked the question, "What rights to researchers have to our top secret information?". In this complex model the answer is not simple, it would depend on many roles in many contexts. Of course this example is extreme, but it highlights that trying to build many barriers in your business can result in a nightmare of administration and confusion amongst users. In the real world what we need is a balance of the two. We need to seek an optimum number of contexts. Too many contexts are unmanageable and too few contexts does not give fine enough granularity. What makes a good context? Good context design derives mainly from how well you understand your business requirements to secure access to confidential information. Some customers I have worked with can tell me exactly the documents they wish to secure and know exactly who should be opening them. However there are some customers who know only of the government regulation that requires them to control access to certain types of information, they don't actually know where the documents are, how they are created or understand exactly who should have access. Therefore you need to know how to ask the business the right questions that lead to information which help you define a context. First ask these questions about a set of documentsWhat is the topic? Who are legitimate contributors on this topic? Who are the authorized readership? If the answer to any one of these is significantly different, then it probably merits a separate context. Remember that sealed documents are inherently secure and as such they cannot leak to your competitors, therefore it is better sealed to a broad context than not sealed at all. Simplicity is key here. Always revert to the first extreme example of a single classification, then work towards essential complexity. If there is any doubt, always prefer fewer contexts. Remember, Oracle IRM allows you to change your mind later on. You can implement a design now and continue to change and refine as you learn how the technology is used. It is easy to go from a simple model to a more complex one, it is much harder to take a complex model that is already embedded in the work practice of users and try to simplify it. It is also wise to take a single use case and address this first with the business. Don't try and tackle many different problems from the outset. Do one, learn from the process, refine it and then take what you have learned into the next use case, refine and continue. Once you have a good grasp of the technology and understand how your business will use it, you can then start rolling out the technology wider across the business. Deciding on the use of roles in the context Once you have decided on that first initial use case and a context to create let's look at the details you need to decide upon. For each context, identify; Administrative rolesBusiness owner, the person who makes decisions about who may or may not see content in this context. This is often the person who wanted to use IRM and drove the business purchase. They are the usually the person with the most at risk when sensitive information is lost. Point of contact, the person who will handle requests for access to content. Sometimes the same as the business owner, sometimes a trusted secretary or administrator. Context administrator, the person who will enact the decisions of the Business Owner. Sometimes the point of contact, sometimes a trusted IT person. Document related rolesContributors, the people who create and edit documents in this context. Reviewers, the people who are involved in reviewing documents but are not trusted to secure information to this classification. This role is not always necessary. (See later discussion on Published-work and Work-in-Progress) Readers, the people who read documents from this context. Some people may have several of the roles above, which is fine. What you are trying to do is understand and define how the business interacts with your sensitive information. These roles obviously map directly to roles available in Oracle IRM. Reviewing the features and security for context roles At this point we have decided on a classification of information, understand what roles people in the business will play when administrating this classification and how they will interact with content. The final piece of the puzzle in getting the information for our first context is to look at the permissions people will have to sealed documents. First think why are you protecting the documents in the first place? It is to prevent the loss of leaking of information to the wrong people. To control the information, making sure that people only access the latest versions of documents. You are not using Oracle IRM to prevent unauthorized people from doing legitimate work. This is an important point, with IRM you can erect many barriers to prevent access to content yet too many restrictions and authorized users will often find ways to circumvent using the technology and end up distributing unprotected originals. Because IRM is a security technology, it is easy to get carried away restricting different groups. However I would highly recommend starting with a simple solution with few restrictions. Ensure that everyone who reasonably needs to read documents can do so from the outset. Remember that with Oracle IRM you can change rights to content whenever you wish and tighten security. Always return to the fact that the greatest value IRM brings is that ONLY authorized users can access secured content, remember that simple "one context for the entire business" model. At the start of the deployment you really need to aim for user acceptance and therefore a simple model is more likely to succeed. As time passes and users understand how IRM works you can start to introduce more restrictions and complexity. Another key aspect to focus on is handling exceptions. If you decide on a context model where engineering can only access engineering information, and sales can only access sales data. Act quickly when a sales manager needs legitimate access to a set of engineering documents. Having a quick and effective process for permitting other people with legitimate needs to obtain appropriate access will be rewarded with acceptance from the user community. These use cases can often be satisfied by integrating IRM with a good Identity & Access Management technology which simplifies the process of assigning users the correct business roles. The big print issue... Printing is often an issue of contention, users love to print but the business wants to ensure sensitive information remains in the controlled digital world. There are many cases of physical document loss causing a business pain, it is often overlooked that IRM can help with this issue by limiting the ability to generate physical copies of digital content. However it can be hard to maintain a balance between security and usability when it comes to printing. Consider the following points when deciding about whether to give print rights. Oracle IRM sealed documents can contain watermarks that expose information about the user, time and location of access and the classification of the document. This information would reside in the printed copy making it easier to trace who printed it. Printed documents are slower to distribute in comparison to their digital counterparts, so time sensitive information in printed format may present a lower risk. Print activity is audited, therefore you can monitor and react to users abusing print rights. Summary In summary it is important to think carefully about the way you create your context model. As you ask the business these questions you may get a variety of different requirements. There may be special projects that require a context just for sensitive information created during the lifetime of the project. There may be a department that requires all information in the group is secured and you might have a few senior executives who wish to use IRM to exchange a small number of highly sensitive documents with a very small number of people. Oracle IRM, with its very flexible context classification system, can support all of these use cases. The trick is to introducing the complexity to deliver them at the right level. In another article i'm working on I will go through some examples of how Oracle IRM might map to existing business use cases. But for now, this article covers all the important questions you need to get your IRM service deployed and successfully protecting your most sensitive information.

  Read the article

• Cisco ASA 5505 - L2TP over IPsec

  - by xraminx

  I have followed this document on cisco site to set up the L2TP over IPsec connection. When I try to establish a VPN to ASA 5505 from my Windows XP, after I click on "connect" button, the "Connecting ...." dialog box appears and after a while I get this error message: Error 800: Unable to establish VPN connection. The VPN server may be unreachable, or security parameters may not be configured properly for this connection. ASA version 7.2(4) ASDM version 5.2(4) Windows XP SP3 Windows XP and ASA 5505 are on the same LAN for test purposes. Edit 1: There are two VLANs defined on the cisco device (the standard setup on cisco ASA5505). - port 0 is on VLAN2, outside; - and ports 1 to 7 on VLAN1, inside. I run a cable from my linksys home router (10.50.10.1) to the cisco ASA5505 router on port 0 (outside). Port 0 have IP 192.168.1.1 used internally by cisco and I have also assigned the external IP 10.50.10.206 to port 0 (outside). I run a cable from Windows XP to Cisco router on port 1 (inside). Port 1 is assigned an IP from Cisco router 192.168.1.2. The Windows XP is also connected to my linksys home router via wireless (10.50.10.141). Edit 2: When I try to establish vpn, the Cisco device real time Log viewer shows 7 entries like this: Severity:5 Date:Sep 15 2009 Time: 14:51:29 SyslogID: 713904 Destination IP = 10.50.10.141, Decription: No crypto map bound to interface... dropping pkt Edit 3: This is the setup on the router right now. Result of the command: "show run" : Saved : ASA Version 7.2(4) ! hostname ciscoasa domain-name default.domain.invalid enable password HGFHGFGHFHGHGFHGF encrypted passwd NMMNMNMNMNMNMN encrypted names name 192.168.1.200 WebServer1 name 10.50.10.206 external-ip-address ! interface Vlan1 nameif inside security-level 100 ip address 192.168.1.1 255.255.255.0 ! interface Vlan2 nameif outside security-level 0 ip address external-ip-address 255.0.0.0 ! interface Vlan3 no nameif security-level 50 no ip address ! interface Ethernet0/0 switchport access vlan 2 ! interface Ethernet0/1 ! interface Ethernet0/2 ! interface Ethernet0/3 ! interface Ethernet0/4 ! interface Ethernet0/5 ! interface Ethernet0/6 ! interface Ethernet0/7 ! ftp mode passive dns server-group DefaultDNS domain-name default.domain.invalid object-group service l2tp udp port-object eq 1701 access-list outside_access_in remark Allow incoming tcp/http access-list outside_access_in extended permit tcp any host WebServer1 eq www access-list outside_access_in extended permit udp any any eq 1701 access-list inside_nat0_outbound extended permit ip any 192.168.1.208 255.255.255.240 access-list inside_cryptomap_1 extended permit ip interface outside interface inside pager lines 24 logging enable logging asdm informational mtu inside 1500 mtu outside 1500 ip local pool PPTP-VPN 192.168.1.210-192.168.1.220 mask 255.255.255.0 icmp unreachable rate-limit 1 burst-size 1 asdm image disk0:/asdm-524.bin no asdm history enable arp timeout 14400 global (outside) 1 interface nat (inside) 0 access-list inside_nat0_outbound nat (inside) 1 0.0.0.0 0.0.0.0 static (inside,outside) tcp interface www WebServer1 www netmask 255.255.255.255 access-group outside_access_in in interface outside timeout xlate 3:00:00 timeout conn 1:00:00 half-closed 0:10:00 udp 0:02:00 icmp 0:00:02 timeout sunrpc 0:10:00 h323 0:05:00 h225 1:00:00 mgcp 0:05:00 mgcp-pat 0:05:00 timeout sip 0:30:00 sip_media 0:02:00 sip-invite 0:03:00 sip-disconnect 0:02:00 timeout sip-provisional-media 0:02:00 uauth 0:05:00 absolute http server enable http 192.168.1.0 255.255.255.0 inside no snmp-server location no snmp-server contact snmp-server enable traps snmp authentication linkup linkdown coldstart crypto ipsec transform-set TRANS_ESP_3DES_SHA esp-3des esp-sha-hmac crypto ipsec transform-set TRANS_ESP_3DES_SHA mode transport crypto ipsec transform-set TRANS_ESP_3DES_MD5 esp-3des esp-md5-hmac crypto ipsec transform-set TRANS_ESP_3DES_MD5 mode transport crypto map outside_map 1 match address inside_cryptomap_1 crypto map outside_map 1 set transform-set TRANS_ESP_3DES_MD5 crypto map outside_map interface inside crypto isakmp enable outside crypto isakmp policy 10 authentication pre-share encryption 3des hash md5 group 2 lifetime 86400 telnet timeout 5 ssh timeout 5 console timeout 0 dhcpd auto_config outside ! dhcpd address 192.168.1.2-192.168.1.33 inside dhcpd enable inside ! group-policy DefaultRAGroup internal group-policy DefaultRAGroup attributes dns-server value 192.168.1.1 vpn-tunnel-protocol IPSec l2tp-ipsec username myusername password FGHFGHFHGFHGFGFHF nt-encrypted tunnel-group DefaultRAGroup general-attributes address-pool PPTP-VPN default-group-policy DefaultRAGroup tunnel-group DefaultRAGroup ipsec-attributes pre-shared-key * tunnel-group DefaultRAGroup ppp-attributes no authentication chap authentication ms-chap-v2 ! ! prompt hostname context Cryptochecksum:a9331e84064f27e6220a8667bf5076c1 : end

  Read the article

• HAProxy: Display a "BADREQ" | BADREQ's by the thousands

  - by GruffTech

  My HAProxy Configuration. #HA-Proxy version 1.3.22 2009/10/14 Copyright 2000-2009 Willy Tarreau <[email protected]> global maxconn 10000 spread-checks 50 user haproxy group haproxy daemon stats socket /tmp/haproxy log localhost local0 log localhost local1 notice defaults mode http maxconn 50000 timeout client 10000 option forwardfor except 127.0.0.1 option httpclose option httplog listen dcaustin 0.0.0.0:80 mode http timeout connect 12000 timeout server 60000 timeout queue 120000 balance roundrobin option httpchk GET /index.html log global option httplog option dontlog-normal server web1 10.10.10.101:80 maxconn 300 check fall 1 server web2 10.10.10.102:80 maxconn 300 check fall 1 server web3 10.10.10.103:80 maxconn 300 check fall 1 server web4 10.10.10.104:80 maxconn 300 check fall 1 listen stats 0.0.0.0:9000 mode http balance log global timeout client 5000 timeout connect 4000 timeout server 30000 stats uri /haproxy HAProxy is running, and the socket is working... adam@dcaustin:/etc/haproxy# echo "show info" | socat stdio /tmp/haproxy Name: HAProxy Version: 1.3.22 Release_date: 2009/10/14 Nbproc: 1 Process_num: 1 Pid: 6320 Uptime: 0d 0h14m58s Uptime_sec: 898 Memmax_MB: 0 Ulimit-n: 20017 Maxsock: 20017 Maxconn: 10000 Maxpipes: 0 CurrConns: 47 PipesUsed: 0 PipesFree: 0 Tasks: 51 Run_queue: 1 node: dcaustin desiption: Errors show nothing from socket... adam@dcaustin:/etc/haproxy# echo "show errors" | socat stdio /tmp/haproxy adam@dcaustin:/etc/haproxy# However... My Error log is exploding with "badrequests" with the Error code cR. cR (according to 1.3 documentation) is The "timeout http-request" stroke before the client sent a full HTTP request. This is sometimes caused by too large TCP MSS values on the client side for PPPoE networks which cannot transport full-sized packets, or by clients sending requests by hand and not typing fast enough, or forgetting to enter the empty line at the end of the request. The HTTP status code is likely a 408 here. Correct on the 408, but we're getting literally thousands of these requests every hour. (This log snippet is an clip for about 10 seconds of time...) Jun 30 11:08:52 localhost haproxy[6320]: 92.22.213.32:26448 [30/Jun/2011:11:08:42.384] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10002 408 212 - - cR-- 35/35/18/0/0 0/0 "<BADREQ>" Jun 30 11:08:54 localhost haproxy[6320]: 71.62.130.24:62818 [30/Jun/2011:11:08:44.457] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10001 408 212 - - cR-- 39/39/16/0/0 0/0 "<BADREQ>" Jun 30 11:08:55 localhost haproxy[6320]: 84.73.75.236:3589 [30/Jun/2011:11:08:45.021] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10008 408 212 - - cR-- 35/35/15/0/0 0/0 "<BADREQ>" Jun 30 11:08:55 localhost haproxy[6320]: 69.39.20.190:49969 [30/Jun/2011:11:08:45.709] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10000 408 212 - - cR-- 37/37/16/0/0 0/0 "<BADREQ>" Jun 30 11:08:56 localhost haproxy[6320]: 2.29.0.9:58772 [30/Jun/2011:11:08:46.846] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10001 408 212 - - cR-- 43/43/22/0/0 0/0 "<BADREQ>" Jun 30 11:08:57 localhost haproxy[6320]: 212.139.250.242:57537 [30/Jun/2011:11:08:47.568] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10000 408 212 - - cR-- 42/42/21/0/0 0/0 "<BADREQ>" Jun 30 11:08:58 localhost haproxy[6320]: 74.79.195.75:55046 [30/Jun/2011:11:08:48.559] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10000 408 212 - - cR-- 46/46/24/0/0 0/0 "<BADREQ>" Jun 30 11:08:58 localhost haproxy[6320]: 74.79.195.75:55044 [30/Jun/2011:11:08:48.554] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10004 408 212 - - cR-- 45/45/24/0/0 0/0 "<BADREQ>" Jun 30 11:08:58 localhost haproxy[6320]: 74.79.195.75:55045 [30/Jun/2011:11:08:48.554] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10005 408 212 - - cR-- 44/44/24/0/0 0/0 "<BADREQ>" Jun 30 11:09:00 localhost haproxy[6320]: 68.197.56.2:52781 [30/Jun/2011:11:08:50.975] dcaustin dcaustin/<NOSRV> -1/-1/-1/-1/10000 408 212 - - cR-- 49/49/28/0/0 0/0 "<BADREQ>" From what I read on google, if i wanted to see what the bad requests are, I can show errors to the socket and it will spit them out. We do run a pretty heavily trafficed website and the percentage of "BADREQS" to normal requests is quite low, but I'd like to be able to get ahold of what that request WAS so I can debug it. stats # pxname,svname,qcur,qmax,scur,smax,slim,stot,bin,bout,dreq,dresp,ereq,econ,eresp,wretr,wredis,status,weight,act,bck,chkfail,chkdown,lastchg,downtime,qlimit,pid,iid,sid,throttle,lbtot,tracked,type,rate,rate_lim,rate_max, dcaustin,FRONTEND,,,64,120,50000,88433,105889100,2553809875,0,0,4641,,,,,OPEN,,,,,,,,,1,1,0,,,,0,45,0,128, dcaustin,web1,0,0,10,28,300,20941,25402112,633143416,,0,,0,3,0,0,UP,1,1,0,0,0,2208,0,,1,1,1,,20941,,2,11,,30, dcaustin,web2,0,0,9,30,300,20941,25026691,641475169,,0,,0,3,0,0,UP,1,1,0,0,0,2208,0,,1,1,2,,20941,,2,11,,30, dcaustin,web3,0,0,10,27,300,20940,30116527,635015040,,0,,0,9,0,0,UP,1,1,0,0,0,2208,0,,1,1,3,,20940,,2,10,,31, dcaustin,web4,0,0,5,28,300,20940,25343770,643209546,,0,,0,8,0,0,UP,1,1,0,0,0,2208,0,,1,1,4,,20940,,2,11,,31, dcaustin,BACKEND,0,0,34,95,50000,83762,105889100,2553809875,0,0,,0,34,0,0,UP,4,4,0,,0,2208,0,,1,1,0,,83762,,1,43,,122, 88500 "Sessions" and 4500 errors. in the last 20 minutes.

  Read the article

• "power limit notification" clobbering on 12G Dell servers with RHEL6

  - by Andrew B

  Server: Poweredge r620 OS: RHEL 6.4 Kernel: 2.6.32-358.18.1.el6.x86_64 I'm experiencing application alarms in my production environment. Critical CPU hungry processes are being starved of resources and causing a processing backlog. The problem is happening on all the 12th Generation Dell servers (r620s) in a recently deployed cluster. As near as I can tell, instances of this happening are matching up to peak CPU utilization, accompanied by massive amounts of "power limit notification" spam in dmesg. An excerpt of one of these events: Nov 7 10:15:15 someserver [.crit] CPU12: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU0: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU6: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU14: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU18: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU2: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU4: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU16: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU0: Package power limit notification (total events = 11) Nov 7 10:15:15 someserver [.crit] CPU6: Package power limit notification (total events = 13) Nov 7 10:15:15 someserver [.crit] CPU14: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU18: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU20: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU8: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU2: Package power limit notification (total events = 12) Nov 7 10:15:15 someserver [.crit] CPU10: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU22: Core power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU4: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU16: Package power limit notification (total events = 13) Nov 7 10:15:15 someserver [.crit] CPU20: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU8: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU10: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU22: Package power limit notification (total events = 14) Nov 7 10:15:15 someserver [.crit] CPU15: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU3: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU1: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU5: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU17: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU13: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU15: Package power limit notification (total events = 375) Nov 7 10:15:15 someserver [.crit] CPU3: Package power limit notification (total events = 374) Nov 7 10:15:15 someserver [.crit] CPU1: Package power limit notification (total events = 376) Nov 7 10:15:15 someserver [.crit] CPU5: Package power limit notification (total events = 376) Nov 7 10:15:15 someserver [.crit] CPU7: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU19: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU17: Package power limit notification (total events = 377) Nov 7 10:15:15 someserver [.crit] CPU9: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU21: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU23: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU11: Core power limit notification (total events = 369) Nov 7 10:15:15 someserver [.crit] CPU13: Package power limit notification (total events = 376) Nov 7 10:15:15 someserver [.crit] CPU7: Package power limit notification (total events = 375) Nov 7 10:15:15 someserver [.crit] CPU19: Package power limit notification (total events = 375) Nov 7 10:15:15 someserver [.crit] CPU9: Package power limit notification (total events = 374) Nov 7 10:15:15 someserver [.crit] CPU21: Package power limit notification (total events = 375) Nov 7 10:15:15 someserver [.crit] CPU23: Package power limit notification (total events = 374) A little Google Fu reveals that this is typically associated with the CPU running hot, or voltage regulation kicking in. I don't think that's what is happening though. Temperature sensors for all servers in the cluster are running fine, Power Cap Policy is disabled in the iDRAC, and my System Profile is set to "Performance" on all of these servers: # omreport chassis biossetup | grep -A10 'System Profile' System Profile Settings ------------------------------------------ System Profile : Performance CPU Power Management : Maximum Performance Memory Frequency : Maximum Performance Turbo Boost : Enabled C1E : Disabled C States : Disabled Monitor/Mwait : Enabled Memory Patrol Scrub : Standard Memory Refresh Rate : 1x Memory Operating Voltage : Auto Collaborative CPU Performance Control : Disabled A Dell mailing list post describes the symptoms almost perfectly. Dell suggested that the author try using the Performance profile, but that didn't help. He ended up applying some settings in Dell's guide for configuring a server for low latency environments and one of those settings (or a combination thereof) seems to have fixed the problem. Kernel.org bug #36182 notes that power-limit interrupt debugging was enabled by default, which is causing performance degradation in scenarios where CPU voltage regulation is kicking in. A RHN KB article (RHN login required) mentions a problem impacting PE r620 and r720 servers not running the Performance profile, and recommends an update to a kernel released two weeks ago. ...Except we are running the Performance profile... Everything I can find online is running me in circles here. What's the heck is going on?

  Read the article

• Why does my MacBook Pro have long ping times over Wi-Fi?

  - by randynov

  I have been having problems connecting with my Wi-Fi. It is weird, the ping times to the router (<30 feet away) seem to surge, often getting over 10 seconds before slowly coming back down. You can see the trend below. I'm on a MacBook Pro and have done the normal stuff (reset the PRAM and SMC, changed wireless channels, etc.). It happens across different routers, so I think it must be my laptop, but I don't know what it could be. The RSSI value hovers around -57, but I've seen the transmit rate flip between 0, 48 and 54. The signal strength is ~60% with 9% noise. Currently, there are 17 other wireless networks in range, but only one in the same channel. 1 - How can I figure out what's going on? 2 - How can I correct the situation? PING 192.168.1.1 (192.168.1.1): 56 data bytes 64 bytes from 192.168.1.1: icmp_seq=0 ttl=254 time=781.107 ms 64 bytes from 192.168.1.1: icmp_seq=1 ttl=254 time=681.551 ms 64 bytes from 192.168.1.1: icmp_seq=2 ttl=254 time=610.001 ms 64 bytes from 192.168.1.1: icmp_seq=3 ttl=254 time=544.915 ms 64 bytes from 192.168.1.1: icmp_seq=4 ttl=254 time=547.622 ms 64 bytes from 192.168.1.1: icmp_seq=5 ttl=254 time=468.914 ms 64 bytes from 192.168.1.1: icmp_seq=6 ttl=254 time=237.368 ms 64 bytes from 192.168.1.1: icmp_seq=7 ttl=254 time=229.902 ms 64 bytes from 192.168.1.1: icmp_seq=8 ttl=254 time=11754.151 ms 64 bytes from 192.168.1.1: icmp_seq=9 ttl=254 time=10753.943 ms 64 bytes from 192.168.1.1: icmp_seq=10 ttl=254 time=9754.428 ms 64 bytes from 192.168.1.1: icmp_seq=11 ttl=254 time=8754.199 ms 64 bytes from 192.168.1.1: icmp_seq=12 ttl=254 time=7754.138 ms 64 bytes from 192.168.1.1: icmp_seq=13 ttl=254 time=6754.159 ms 64 bytes from 192.168.1.1: icmp_seq=14 ttl=254 time=5753.991 ms 64 bytes from 192.168.1.1: icmp_seq=15 ttl=254 time=4754.068 ms 64 bytes from 192.168.1.1: icmp_seq=16 ttl=254 time=3753.930 ms 64 bytes from 192.168.1.1: icmp_seq=17 ttl=254 time=2753.768 ms 64 bytes from 192.168.1.1: icmp_seq=18 ttl=254 time=1753.866 ms 64 bytes from 192.168.1.1: icmp_seq=19 ttl=254 time=753.592 ms 64 bytes from 192.168.1.1: icmp_seq=20 ttl=254 time=517.315 ms 64 bytes from 192.168.1.1: icmp_seq=37 ttl=254 time=1.315 ms 64 bytes from 192.168.1.1: icmp_seq=38 ttl=254 time=1.035 ms 64 bytes from 192.168.1.1: icmp_seq=39 ttl=254 time=4.597 ms 64 bytes from 192.168.1.1: icmp_seq=21 ttl=254 time=18010.681 ms 64 bytes from 192.168.1.1: icmp_seq=22 ttl=254 time=17010.449 ms 64 bytes from 192.168.1.1: icmp_seq=23 ttl=254 time=16010.430 ms 64 bytes from 192.168.1.1: icmp_seq=24 ttl=254 time=15010.540 ms 64 bytes from 192.168.1.1: icmp_seq=25 ttl=254 time=14010.450 ms 64 bytes from 192.168.1.1: icmp_seq=26 ttl=254 time=13010.175 ms 64 bytes from 192.168.1.1: icmp_seq=27 ttl=254 time=12010.282 ms 64 bytes from 192.168.1.1: icmp_seq=28 ttl=254 time=11010.265 ms 64 bytes from 192.168.1.1: icmp_seq=29 ttl=254 time=10010.285 ms 64 bytes from 192.168.1.1: icmp_seq=30 ttl=254 time=9010.235 ms 64 bytes from 192.168.1.1: icmp_seq=31 ttl=254 time=8010.399 ms 64 bytes from 192.168.1.1: icmp_seq=32 ttl=254 time=7010.144 ms 64 bytes from 192.168.1.1: icmp_seq=33 ttl=254 time=6010.113 ms 64 bytes from 192.168.1.1: icmp_seq=34 ttl=254 time=5010.025 ms 64 bytes from 192.168.1.1: icmp_seq=35 ttl=254 time=4009.966 ms 64 bytes from 192.168.1.1: icmp_seq=36 ttl=254 time=3009.825 ms 64 bytes from 192.168.1.1: icmp_seq=40 ttl=254 time=16000.676 ms 64 bytes from 192.168.1.1: icmp_seq=41 ttl=254 time=15000.477 ms 64 bytes from 192.168.1.1: icmp_seq=42 ttl=254 time=14000.388 ms 64 bytes from 192.168.1.1: icmp_seq=43 ttl=254 time=13000.549 ms 64 bytes from 192.168.1.1: icmp_seq=44 ttl=254 time=12000.469 ms 64 bytes from 192.168.1.1: icmp_seq=45 ttl=254 time=11000.332 ms 64 bytes from 192.168.1.1: icmp_seq=46 ttl=254 time=10000.339 ms 64 bytes from 192.168.1.1: icmp_seq=47 ttl=254 time=9000.338 ms 64 bytes from 192.168.1.1: icmp_seq=48 ttl=254 time=8000.198 ms 64 bytes from 192.168.1.1: icmp_seq=49 ttl=254 time=7000.388 ms 64 bytes from 192.168.1.1: icmp_seq=50 ttl=254 time=6000.217 ms 64 bytes from 192.168.1.1: icmp_seq=51 ttl=254 time=5000.084 ms 64 bytes from 192.168.1.1: icmp_seq=52 ttl=254 time=3999.920 ms 64 bytes from 192.168.1.1: icmp_seq=53 ttl=254 time=3000.010 ms 64 bytes from 192.168.1.1: icmp_seq=54 ttl=254 time=1999.832 ms 64 bytes from 192.168.1.1: icmp_seq=55 ttl=254 time=1000.072 ms 64 bytes from 192.168.1.1: icmp_seq=58 ttl=254 time=1.125 ms 64 bytes from 192.168.1.1: icmp_seq=59 ttl=254 time=1.070 ms 64 bytes from 192.168.1.1: icmp_seq=60 ttl=254 time=2.515 ms

  Read the article

• Cisco PIX 8.0.4, static address mapping not working?

  - by Bill

  upgrading a working Pix running 5.3.1 to 8.0.4. The memory/IOS upgrade went fine, but the 8.0.4 configuration is not quite working 100%. The 5.3.1 config on which it was based is working fine. Basically, I have three networks (inside, outside, dmz) with some addresses on the dmz statically mapped to outside addresses. The problem seems to be that those addresses can't send or receive traffic from the outside (Internet.) Stuff on the DMZ that does not have a static mapping seems to work fine. So, basically: Inside - outside: works Inside - DMZ: works DMZ - inside: works, where the rules allow it DMZ (non-static) - outside: works But: DMZ (static) - outside: fails Outside - DMZ: fails (So, say, udp 1194 traffic to .102, http to .104) I suspect there's something I'm missing with the nat/global section of the config, but can't for the life of me figure out what. Help, anyone? The complete configuration is below. Thanks for any thoughts! ! PIX Version 8.0(4) ! hostname firewall domain-name asasdkpaskdspakdpoak.com enable password xxxxxxxx encrypted passwd xxxxxxxx encrypted names ! interface Ethernet0 nameif outside security-level 0 ip address XX.XX.XX.100 255.255.255.224 ! interface Ethernet1 nameif inside security-level 100 ip address 192.168.68.1 255.255.255.0 ! interface Ethernet2 nameif dmz security-level 10 ip address 192.168.69.1 255.255.255.0 ! boot system flash:/image.bin ftp mode passive dns server-group DefaultDNS domain-name asasdkpaskdspakdpoak.com access-list acl_out extended permit udp any host XX.XX.XX.102 eq 1194 access-list acl_out extended permit tcp any host XX.XX.XX.104 eq www access-list acl_dmz extended permit tcp host 192.168.69.10 host 192.168.68.17 eq ssh access-list acl_dmz extended permit tcp 10.71.83.0 255.255.255.0 192.168.68.0 255.255.255.0 eq ssh access-list acl_dmz extended permit tcp 10.71.83.0 255.255.255.0 192.168.68.0 255.255.255.0 eq 5901 access-list acl_dmz extended permit udp host 192.168.69.103 any eq ntp access-list acl_dmz extended permit udp host 192.168.69.103 any eq domain access-list acl_dmz extended permit tcp host 192.168.69.103 any eq www access-list acl_dmz extended permit tcp host 192.168.69.100 host 192.168.68.101 eq 3306 access-list acl_dmz extended permit tcp host 192.168.69.100 host 192.168.68.102 eq 3306 access-list acl_dmz extended permit tcp host 192.168.69.101 host 192.168.68.101 eq 3306 access-list acl_dmz extended permit tcp host 192.168.69.101 host 192.168.68.102 eq 3306 access-list acl_dmz extended permit tcp 10.71.83.0 255.255.255.0 host 192.168.68.101 eq 3306 access-list acl_dmz extended permit tcp 10.71.83.0 255.255.255.0 host 192.168.68.102 eq 3306 access-list acl_dmz extended permit tcp host 192.168.69.104 host 192.168.68.101 eq 3306 access-list acl_dmz extended permit tcp host 192.168.69.104 host 192.168.68.102 eq 3306 access-list acl_dmz extended permit tcp 10.71.83.0 255.255.255.0 host 192.168.69.104 eq 8080 access-list acl_dmz extended permit tcp 10.71.83.0 255.255.255.0 host 192.168.69.104 eq 8099 access-list acl_dmz extended permit tcp host 192.168.69.105 any eq www access-list acl_dmz extended permit tcp host 192.168.69.103 any eq smtp access-list acl_dmz extended permit tcp host 192.168.69.105 host 192.168.68.103 eq ssh access-list acl_dmz extended permit tcp host 192.168.69.104 any eq www access-list acl_dmz extended permit tcp host 192.168.69.100 any eq www access-list acl_dmz extended permit tcp host 192.168.69.100 any eq https pager lines 24 mtu outside 1500 mtu inside 1500 mtu dmz 1500 icmp unreachable rate-limit 1 burst-size 1 no asdm history enable arp timeout 14400 global (outside) 1 interface nat (inside) 1 0.0.0.0 0.0.0.0 nat (dmz) 1 0.0.0.0 0.0.0.0 static (dmz,outside) XX.XX.XX.103 192.168.69.11 netmask 255.255.255.255 static (inside,dmz) 192.168.68.17 192.168.68.17 netmask 255.255.255.255 static (inside,dmz) 192.168.68.100 192.168.68.100 netmask 255.255.255.255 static (inside,dmz) 192.168.68.101 192.168.68.101 netmask 255.255.255.255 static (inside,dmz) 192.168.68.102 192.168.68.102 netmask 255.255.255.255 static (inside,dmz) 192.168.68.103 192.168.68.103 netmask 255.255.255.255 static (dmz,outside) XX.XX.XX.104 192.168.69.100 netmask 255.255.255.255 static (dmz,outside) XX.XX.XX.105 192.168.69.105 netmask 255.255.255.255 static (dmz,outside) XX.XX.XX.102 192.168.69.10 netmask 255.255.255.255 access-group acl_out in interface outside access-group acl_dmz in interface dmz route outside 0.0.0.0 0.0.0.0 XX.XX.XX.97 1 route dmz 10.71.83.0 255.255.255.0 192.168.69.10 1 timeout xlate 3:00:00 timeout conn 1:00:00 half-closed 0:10:00 udp 0:02:00 icmp 0:00:02 timeout sunrpc 0:10:00 h323 0:05:00 h225 1:00:00 mgcp 0:05:00 mgcp-pat 0:05:00 timeout sip 0:30:00 sip_media 0:02:00 sip-invite 0:03:00 sip-disconnect 0:02:00 timeout sip-provisional-media 0:02:00 uauth 0:05:00 absolute dynamic-access-policy-record DfltAccessPolicy no snmp-server location no snmp-server contact snmp-server enable traps snmp authentication linkup linkdown coldstart crypto ipsec security-association lifetime seconds 28800 crypto ipsec security-association lifetime kilobytes 4608000 telnet 192.168.68.17 255.255.255.255 inside telnet timeout 5 ssh timeout 5 console timeout 0 threat-detection basic-threat threat-detection statistics access-list no threat-detection statistics tcp-intercept ! class-map inspection_default match default-inspection-traffic ! ! policy-map type inspect dns preset_dns_map parameters message-length maximum 512 policy-map global_policy class inspection_default inspect dns preset_dns_map inspect ftp inspect h323 h225 inspect h323 ras inspect netbios inspect rsh inspect rtsp inspect skinny inspect esmtp inspect sqlnet inspect sunrpc inspect tftp inspect sip inspect xdmcp ! service-policy global_policy global prompt hostname context Cryptochecksum:2d1bb2dee2d7a3e45db63a489102d7de

  Read the article

• Why does my macbook pro have long ping times over wifi?

  - by randynov

  I have been having problems connecting with my wifi. It is weird, the ping times to the router (<30 feet away) seem to surge, often getting over 10s before slowly coming back down. You can see the trend below. I'm on a macbook pro and have done the normal stuff (reset the pram and smc, changed wireless channels, etc.). It happens across different routers, so I think it must be my laptop, but I don't know what it could be. The RSSI value hovers around -57, but I've seen the transmit rate flip between 0, 48 & 54. The signal strength is ~60% with 9% noise. Currently, there are 17 other wireless networks in range, but only one in the same channel. 1 - How can I figure out what's going on? 2 - How can I correct the situation? TIA! Randall PING 192.168.1.1 (192.168.1.1): 56 data bytes 64 bytes from 192.168.1.1: icmp_seq=0 ttl=254 time=781.107 ms 64 bytes from 192.168.1.1: icmp_seq=1 ttl=254 time=681.551 ms 64 bytes from 192.168.1.1: icmp_seq=2 ttl=254 time=610.001 ms 64 bytes from 192.168.1.1: icmp_seq=3 ttl=254 time=544.915 ms 64 bytes from 192.168.1.1: icmp_seq=4 ttl=254 time=547.622 ms 64 bytes from 192.168.1.1: icmp_seq=5 ttl=254 time=468.914 ms 64 bytes from 192.168.1.1: icmp_seq=6 ttl=254 time=237.368 ms 64 bytes from 192.168.1.1: icmp_seq=7 ttl=254 time=229.902 ms 64 bytes from 192.168.1.1: icmp_seq=8 ttl=254 time=11754.151 ms 64 bytes from 192.168.1.1: icmp_seq=9 ttl=254 time=10753.943 ms 64 bytes from 192.168.1.1: icmp_seq=10 ttl=254 time=9754.428 ms 64 bytes from 192.168.1.1: icmp_seq=11 ttl=254 time=8754.199 ms 64 bytes from 192.168.1.1: icmp_seq=12 ttl=254 time=7754.138 ms 64 bytes from 192.168.1.1: icmp_seq=13 ttl=254 time=6754.159 ms 64 bytes from 192.168.1.1: icmp_seq=14 ttl=254 time=5753.991 ms 64 bytes from 192.168.1.1: icmp_seq=15 ttl=254 time=4754.068 ms 64 bytes from 192.168.1.1: icmp_seq=16 ttl=254 time=3753.930 ms 64 bytes from 192.168.1.1: icmp_seq=17 ttl=254 time=2753.768 ms 64 bytes from 192.168.1.1: icmp_seq=18 ttl=254 time=1753.866 ms 64 bytes from 192.168.1.1: icmp_seq=19 ttl=254 time=753.592 ms 64 bytes from 192.168.1.1: icmp_seq=20 ttl=254 time=517.315 ms 64 bytes from 192.168.1.1: icmp_seq=37 ttl=254 time=1.315 ms 64 bytes from 192.168.1.1: icmp_seq=38 ttl=254 time=1.035 ms 64 bytes from 192.168.1.1: icmp_seq=39 ttl=254 time=4.597 ms 64 bytes from 192.168.1.1: icmp_seq=21 ttl=254 time=18010.681 ms 64 bytes from 192.168.1.1: icmp_seq=22 ttl=254 time=17010.449 ms 64 bytes from 192.168.1.1: icmp_seq=23 ttl=254 time=16010.430 ms 64 bytes from 192.168.1.1: icmp_seq=24 ttl=254 time=15010.540 ms 64 bytes from 192.168.1.1: icmp_seq=25 ttl=254 time=14010.450 ms 64 bytes from 192.168.1.1: icmp_seq=26 ttl=254 time=13010.175 ms 64 bytes from 192.168.1.1: icmp_seq=27 ttl=254 time=12010.282 ms 64 bytes from 192.168.1.1: icmp_seq=28 ttl=254 time=11010.265 ms 64 bytes from 192.168.1.1: icmp_seq=29 ttl=254 time=10010.285 ms 64 bytes from 192.168.1.1: icmp_seq=30 ttl=254 time=9010.235 ms 64 bytes from 192.168.1.1: icmp_seq=31 ttl=254 time=8010.399 ms 64 bytes from 192.168.1.1: icmp_seq=32 ttl=254 time=7010.144 ms 64 bytes from 192.168.1.1: icmp_seq=33 ttl=254 time=6010.113 ms 64 bytes from 192.168.1.1: icmp_seq=34 ttl=254 time=5010.025 ms 64 bytes from 192.168.1.1: icmp_seq=35 ttl=254 time=4009.966 ms 64 bytes from 192.168.1.1: icmp_seq=36 ttl=254 time=3009.825 ms 64 bytes from 192.168.1.1: icmp_seq=40 ttl=254 time=16000.676 ms 64 bytes from 192.168.1.1: icmp_seq=41 ttl=254 time=15000.477 ms 64 bytes from 192.168.1.1: icmp_seq=42 ttl=254 time=14000.388 ms 64 bytes from 192.168.1.1: icmp_seq=43 ttl=254 time=13000.549 ms 64 bytes from 192.168.1.1: icmp_seq=44 ttl=254 time=12000.469 ms 64 bytes from 192.168.1.1: icmp_seq=45 ttl=254 time=11000.332 ms 64 bytes from 192.168.1.1: icmp_seq=46 ttl=254 time=10000.339 ms 64 bytes from 192.168.1.1: icmp_seq=47 ttl=254 time=9000.338 ms 64 bytes from 192.168.1.1: icmp_seq=48 ttl=254 time=8000.198 ms 64 bytes from 192.168.1.1: icmp_seq=49 ttl=254 time=7000.388 ms 64 bytes from 192.168.1.1: icmp_seq=50 ttl=254 time=6000.217 ms 64 bytes from 192.168.1.1: icmp_seq=51 ttl=254 time=5000.084 ms 64 bytes from 192.168.1.1: icmp_seq=52 ttl=254 time=3999.920 ms 64 bytes from 192.168.1.1: icmp_seq=53 ttl=254 time=3000.010 ms 64 bytes from 192.168.1.1: icmp_seq=54 ttl=254 time=1999.832 ms 64 bytes from 192.168.1.1: icmp_seq=55 ttl=254 time=1000.072 ms 64 bytes from 192.168.1.1: icmp_seq=58 ttl=254 time=1.125 ms 64 bytes from 192.168.1.1: icmp_seq=59 ttl=254 time=1.070 ms 64 bytes from 192.168.1.1: icmp_seq=60 ttl=254 time=2.515 ms

  Read the article

< Previous Page | 105 106 107 108 109 110 111 112 113 114 115 116  | Next Page >