Search Results

Search found 143 results on 6 pages for 'perf'.

Page 4/6 | < Previous Page | 1 2 3 4 5 6  | Next Page >

  • Mirror a RAID0 volume

    - by Ghostrider
    I have two SSD running in RAID0. The capacity and speed are just great. I use Windows Home Server to do incremental daily backups. This is fine and well and I've successfully restored from these backups. However. When one of the disks physically died. I was stuck without a working system until the replacement arrives so that I can restore the array from backup. WHS restoration takes about 5 hours which basically means that I'm losing entire day for the process. Is it possible to set up kind of a recovery volume for the RAID array? Use a single mechanical HDD that would be updated with the exact clone of the RAID array on a daily basis. This way if the array goes offline for some reason, I can just boot from the mechanical HDD, lose some perf but will still be able to work. The machine in question runs Windows 7. Creating RAID01 is not an option because of the high price of the SSD and the fact that it still doesn't protect against failure of RAID controller. Is there any way it can be set up?

    Read the article

  • How to monitor process hogging CPU on a remote server

    - by sergeb
    I have a remotely hosted (virtual, VMware) dedicated server (Windows 2008 Server Web edition w/ SP1) that I can only connect to over Remote Desktop. Lately, a process hogs CPU for ~40 minutes most every day (at a random hour) and brings all web sites on the server down. While this is going on I also cannot connect using Remote Desktop to investigate on what is that process... Promptly after 40 min I can RD and the first thing I see on the Perf Monitor is that there was something topping the CPU at 100% and stops just before I'm able to RD... I'm aware of the beginning and end of this for I have monitors setup that email me up/down status of the web sites but I'm locked out while this is happening - can't RD to the server until it's over (and too late to see the Task Manager/Process Explorer picture). What is the best way/tool to setup on the server to continuously monitor all processes so when this happens I login and "replay" it to find the process causing this trouble? (I have no control over the virtual/VMware setup for it is hosted by a 3rd-party but I have most full control over my dedicated machine) Thanks in advance!

    Read the article

  • Poor Write Performance in VM inside Proxmox PVE 2.0

    - by sorsenne
    I am running a PVE 2.0 on a decent Hardware (2 SATA HDDs as RAID1, 12GB RAM, i7 CPU) but the I/O Performance is very poor inside the VM (Ubuntu 11.10 Server). The very same VM was copied to another Server running simply Ubuntu Server with KVM and had better I/O Perf. this is how the HDD is shown in the Guest: ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300) ata1.00: ATA-8: ST3000DM001-9YN166, CC49, max UDMA/133 ata1.00: 5860533168 sectors, multi 16: LBA48 NCQ (depth 31/32), AA ata1.00: configured for UDMA/133 scsi 0:0:0:0: Direct-Access ATA ST3000DM001-9YN1 CC49 PQ: 0 ANSI: 5 sd 0:0:0:0: [sda] 5860533168 512-byte logical blocks: (3.00 TB/2.72 TiB) sd 0:0:0:0: [sda] 4096-byte physical blocks sd 0:0:0:0: [sda] Write Protect is off sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00 sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA I tested with DD: $ dd bs=1M count=128 if=/dev/zero of=test conv=fdatasync 128+0 records in 128+0 records out 134217728 bytes (134 MB) copied, 19.2222 s, 7.0 MB/s on the Host, this same Test will result with 156 MB/s in average. PS: I am using VirtIO and see no error in dmesg.

    Read the article

  • Live Mesh starts exactly once on fresh Win7 Ultimate installation

    - by Reb.Cabin
    I did a fresh install of Windows 7 Ulimate 64-bit on a formatted drive on a refurbed Lenovo PC, applied all 102 (!) windows updates, windows seems to be working fine. No quirks installing, no apps, no junkware, just straight, legal, Win7 Ultimate right from an unopened 2009 Microsoft box. Ok, breathe sigh -- Install Live Mesh (no messenger, no mail, no writer, no photo, none of the rest of the Windows Live freeware). Set up my shares, let it run overnight. watch MOE.exe in the Task-Manager perf pane to make sure it's all settled down. reboot. Ok, check that MOE is running and files are getting updated properly from other machines in the mesh. Great. HOWEVER -- when I try to launch Windows Live Mesh app from the Start jewel, I get a brief hourglass, then nothing. Reboot. Same story. result -- the shares I already posted seem to be synching properly, but I can't run the app, so I can't add and delete shares. The background process MOE seems to run, but I can't get the app going. btw, the reason I did this fresh install is I had exactly the same experience running Vista, so I wiped the machine hoping it would solve this nasty problem. Imagine my surprise! Will be grateful for clues, advice, etc, please & thanks!

    Read the article

  • DNS-Based Environment Determination

    - by zvolkov
    Found the following here. The questions is: where can I find more details on how exactly implement this on Windows? Any guide or how-to anybody? Or maybe you can provide your invaluable suggestions? Specifically, how do I make so that "all QA servers would first resolve entries in qa.example.com first and then if that lookup failed they would try example.com" (I'm a dev, not a DNS specialist, but our IT Support has refused to help on this:() Use DNS Based Environment Determination for your servers. Do this by initially splitting your top level domain into a number of sub domains depending on their function, and then creating DNS Service Names in each of the sub domains pointing to the relevant server for that service. Based on the list above we would then have: * clientdb.prod.example.com for Production * clientdb.perf.example.com for Performance Testing * clientdb.qa.example.com for QA * clientdb.dev.example.com for Development Servers then resolve entries in their relevant sub domain by function. That is, all QA servers would first resolve entries in qa.example.com first and then if that lookup failed they would try example.com. This allows you to have a single configuration entry for your client database hostname (clientdb) that would resolve correctly in all environments. This technique has the added advantage of still having global services defined in a common top level domain. This seems to be related to Providing "split horizon" DNS service. Reading that, I see that I will probably need separate DNS Server for each environment. Is this true or does Windows support some form of "tagging" the records to be visible depending on the requestor's IP?

    Read the article

  • Ruby installed on Ubuntu 10.10 slow on one machine but not other

    - by Aaron Jensen
    I have a machine that was provisioned several months ago. RVM was used to install ruby 1.9.3-p125 as well as 1.9.3-p125-perf. When I compared raw ruby performance to another identical machine the older machine smoked them. For example: ================================================================================ With in-block needle calculation ================================================================================ Rehearsal ---------------------------------------------- detect 3.790000 0.000000 3.790000 ( 3.800895) each 2.410000 0.000000 2.410000 ( 2.420860) any 3.960000 0.000000 3.960000 ( 3.972099) include 1.440000 0.000000 1.440000 ( 1.442862) ------------------------------------ total: 11.600000sec vs ================================================================================ With in-block needle calculation ================================================================================ Rehearsal ---------------------------------------------- detect 10.740000 0.000000 10.740000 ( 10.769366) each 6.080000 0.010000 6.090000 ( 6.106323) any 10.600000 0.000000 10.600000 ( 10.641606) include 4.160000 0.000000 4.160000 ( 4.171530) ------------------------------------ total: 31.590000sec I attempted to reinstall 1.9.3-p125 with rvm on the fast machine and that ruby is now slow. It's as if something changed in RVM, or I installed some package that made compiled versions of ruby perform significantly worse. I know this is a tough question to answer, but what things should I look into in order to track down why the performance has suffered so much? edit I just attempted to install with ruby-build and the version installed was fast. Something rvm is doing to build it in my environment is slow.

    Read the article

  • Windows Small Business System 2003. SQL timeout in Server Performance Report

    - by tetranz
    I'm the volunteer IT admin at a small school. We have SBS 2003 with about ten desktops. The server performance report is emailed to me daily. It is setup with a wizard in the Monitoring and Performance part of the "Server Management" console. It often fails with a "The page cannot be displayed" error. The event log shows Event Type: Error Event Source: ServerStatusReports Event Category: None Event ID: 1 Date: 1/16/2011 Time: 6:03:14 AM User: N/A Computer: ALPHA Description: Server Status Report: URL: http://localhost/monitoring/perf.aspx?reportMode=1&allHours=1 Error Message: Timeout expired. The timeout period elapsed prior to completion of the operation or the server is not responding. Stack Trace: at System.Data.SqlClient.SqlConnection.OnError(SqlException exception, TdsParserState state) at System.Data.SqlClient.SqlInternalConnection.OnError(SqlException exception, TdsParserState state) at System.Data.SqlClient.TdsParser.ThrowExceptionAndWarning() at System.Data.SqlClient.TdsParser.ReadNetlib(Int32 bytesExpected) [plus lots more stack trace] This has been happening for years :) I've never really solved it. It seems to be related to WSUS. When it happens, I run the Update Services "Server Cleanup Wizard". That takes a long time to run. If I haven't run it for a while it can take 10 hours. I also run the WsusDBMaintenance.sql script (from TechNet I think) which reindexes the database etc. Those two things seem to get it working again for a while. Recently the "while" has become a couple of weeks. My searching online has revealed lots of people having this problem but no real solution. Does anyone have any good ideas about this? I have to wonder if something in the WSUS SQL schema is not indexed properly. The time that the server cleanup wizard takes seems ridiculous. Thanks

    Read the article

  • Some VS 2010 RC Updates (including patches for Intellisense and Web Designer fixes)

    - by ScottGu
    [In addition to blogging, I am also now using Twitter for quick updates and to share links. Follow me at: twitter.com/scottgu] We are continuing to make progress on shipping Visual Studio 2010.  I’d like to say a big thank you to everyone who has downloaded and tried out the VS 2010 Release Candidate, and especially to those who have sent us feedback or reported issues with it. This data has been invaluable in helping us find and fix remaining bugs before we ship the final release. Last month I blogged about a patch we released for the VS 2010 RC that fixed a bad intellisense crash issue.  This past week we released two additional patches that you can download and apply to the VS 2010 RC to immediately fix two other common issues we’ve seen people run into: Patch that fixes crashes with Tooltip invocation and when hovering over identifiers The Visual Studio team recently released a second patch that fixes some crashes we’ve seen when tooltips are displayed – most commonly when hovering over an identifier to view a QuickInfo tooltip. You can learn more about this issue from this blog post, and download and apply the patch here. Patch that fixes issues with the Web Forms designer not correctly adding controls to the auto-generated designer files The Visual Web Developer team recently released a patch that fixes issues where web controls are not correctly added to the .designer.cs file associated with the .aspx file – which means they can’t be programmed against in the code-behind file.  This issue is most commonly described as “controls are not being recognized in the code-behind” or “editing existing .aspx files regenerates the .aspx.designer.(vb or cs) file and controls are now missing” or “I can’t embed controls within the Ajax Control Toolkit TabContainer or the <asp:createuserwizard> control”. You can learn more about the issue here, and download the patch that fixes it here. Common Cause of Intellisense and IDE sluggishness on Windows XP, Vista, Win Server 2003/2008 systems Over the last few months we’ve occasionally seen reports of people seeing tremendous slowness when typing and using intellisense within VS 2010 despite running on decent machines.  It took us awhile to track down the cause – but we have found that the common culprit seems to be that these machines don’t have the latest versions of the UIA (Windows Automation) component installed. UIA 3 ships with Windows 7, and is a recommended Windows Update patch on XP and Vista (which is why we didn’t see the problem in our tests – since our machines are patched with all recommended updates).  Many systems (especially on XP) don’t automatically install recommended updates, though, and are running with older versions of UIA. This can cause significant performance slow-downs within the VS 2010 editor when large lists are displayed (for example: with intellisense). If you are running on Windows XP, Vista, or Windows Server 2003 or 2008 and are seeing any performance issues with the editor or IDE, please install the free UIA 3 update that can be downloaded from this page.  If you scroll down the page you’ll find direct links to versions for each OS. Note that we are making improvements to the final release of VS 2010 so that we don’t have big perf issues when UIA 3 isn’t installed – and we are also adding a message within the IDE that will warn you if you don’t have UIA 3 installed and accessibility is activated. Improved Text Rendering with WPF 4 and VS 2010 We recently made some nice changes to WPF 4 which improve the text clarity and text crispness over what was in the VS 2010/.NET 4 Release Candidate.  In particular these changes improve scenarios where you have a dark background with light text. You can learn more about these improvements in this WPF Team blog post.  These changes will be in the final release of VS 2010 and .NET 4. Hope this helps, Scott

    Read the article

  • Do NOT Change "Copy Local” project references to false, unless understand subsequences.

    - by Michael Freidgeim
    To optimize performance of visual studio build I've found multiple recommendations to change CopyLocal property for dependent dlls to false,e.g. From http://stackoverflow.com/questions/690033/best-practices-for-large-solutions-in-visual-studio-2008 CopyLocal? For sure turn this offhttp://stackoverflow.com/questions/280751/what-is-the-best-practice-for-copy-local-and-with-project-referencesAlways set the Copy Local property to false and enforce this via a custom msbuild stephttp://codebetter.com/patricksmacchia/2007/06/20/benefit-from-the-c-and-vb-net-compilers-perf/BenefitBenefitMy advice is to always set ‘Copy Local’ to falseSome time ago we've tried to change the setting to false, and found that it causes problem for deployment of top-level projects.Recently I've followed the suggestion and changed the settings for middle-level projects. It didn't cause immediate issues, but I was warned by Readify Consultant Colin Savage about possible errors during deploymentsI haven't undone the changes immediately and we found a few issues during testing.There are many scenarios, when you need to have Copy Local’ left to True.The concerns are highlighted in some stack overflow answers, but they have small number of votes.Top-level projects:  set copy local = true.First of all, it doesn't work correctly for top-level projects, i.e. executables or web sites.As pointed in the answer http://stackoverflow.com/a/6529461/52277for all the references in the one at the top set copy local = true.Alternatively you have to change output directory as it's described in http://www.simple-talk.com/dotnet/.net-framework/partitioning-your-code-base-through-.net-assemblies-and-visual-studio-projects/If you set ‘ Copy Local = false’, VS will, unless you tell it otherwise, place each assembly alone in its own .\bin\Debugdirectory. Because of this, you will need to configure VS to place assemblies together in the same directory. To do so, for each VS project, go to VS > Project Properties > Build tab > Output path, and set the Ouput path to ..\bin\Debugfor debug configuration, and ..\bin\Release for release configuration.Second-level  dependencies:  set copy local = true.Another example when copylocal =false fails on run-time, is when top level assembly doesn't directly referenced one of indirect dependencies.E..g. Top-level assembly A has reference to assembly B with copylocal =true, but assembly B has reference to assembly C with copylocal =false. Most likely assembly C will be missing on runtime and will cause errors E.g. http://stackoverflow.com/questions/602765/when-should-copy-local-be-set-to-true-and-when-should-it-not?lq=1Copy local is important for deployment scenarios and tools. As a general rule you should use CopyLocal=True and http://stackoverflow.com/questions/602765/when-should-copy-local-be-set-to-true-and-when-should-it-not?lq=1 Unfortunately there are some quirks and CopyLocal won't necessary work as expected for assembly references in secondary assemblies structured as shown below.MainApp.exe MyLibrary.dll ThirdPartyLibrary.dll (if in the GAC CopyLocal won't copy to MainApp bin folder)This makes xcopy deployments difficult . .Reflection called DLLs  dependencies:  set copy local = true.E.g user can see error "ISystem.Reflection.ReflectionTypeLoadException: Unable to load one or more of the requested types. Retrieve the LoaderExceptions property for more information."The fix for the issue is recommended in http://stackoverflow.com/a/6200173/52277"I solved this issue by setting the Copy Local attribute of my project's references to true."In general, the problems with investigation of deployment issues may overweight the benefits of reduced build time. Setting the Copy Local to false without considering deployment issues is not a good idea.

    Read the article

  • WPF - LayoutUpdated event firing repeatedly

    - by Drew Noakes
    I've been adding a bit of animation to my WPF application. Thanks to Dan Crevier's unique solution to animating the children of a panel combined with the awesome WPF Penner animations it turned out to be fairly straightforward to make one of my controls look great and have its children move about with some nice animation. Unfortunately this all comes with a performance overhead. I'm happy to have the performance hit when items are added/removed or the control is resized, but it seems that this perf hit occurs consistently throughout the application's lifetime, even when items are completely static. The PanelLayoutAnimator class uses an attached property to hook the UIElement.LayoutUpdated.aspx) event. When this event fires, render transforms are animated to cause the children to glide to their new positions. Unfortunately it seems that the LayoutUpdated event fires every second or so, even when nothing is happening in the application (at least I don't think my code's doing anything -- the app doesn't have focus and the mouse is steady.) As the reason for the event is not immediately apparent to the event handler, all children of the control have to be reevaluated. This event is being called about once a second when idle. The frequency increases when actually using the app. So my question is, how can I improve the performance here? Any answer that assists would be appreciated, but I'm currently stuck on these sub-questions: What causes the LayoutUpdated event to fire so frequently? Is this supposed to happen, and if not, how can I find out why it's firing and curtail it? Is there a more convenient way within the handler to know whether something has happened that might have moved children? If so, I could bail out early and avoid the overhead of looping each child. For now I will work around this issue by disabling animation when there are more than N children in the panel.

    Read the article

  • Find the "largest" dense sub matrix in a large sparse matrix

    - by BCS
    Given a large sparse matrix (say 10k+ by 1M+) I need to find a subset, not necessarily continuous, of the rows and columns that form a dense matrix (all non-zero elements). I want this sub matrix to be as large as possible (not the largest sum, but the largest number of elements) within some aspect ratio constraints. Are there any known exact or aproxamate solutions to this problem? A quick scan on Google seems to give a lot of close-but-not-exactly results. What terms should I be looking for? edit: Just to clarify; the sub matrix need not be continuous. In fact the row and column order is completely arbitrary so adjacency is completely irrelevant. A thought based on Chad Okere's idea Order the rows from largest count to smallest count (not necessary but might help perf) Select two rows that have a "large" overlap Add all other rows that won't reduce the overlap Record that set Add whatever row reduces the overlap by the least Repeat at #3 until the result gets to small Start over at #2 with a different starting pair Continue until you decide the result is good enough

    Read the article

  • Is Rails Metal (& Rack) a good way to implement a high traffic web service api?

    - by Greg
    I am working on a very typical web application. The main component of the user experience is a widget that a site owner would install on their front page. Every time their front page loads, the widget talks to our server and displays some of the data that returns. So there are two components to this web application: the front end UI that the site owner uses to configure their widget the back end component that responds to the widget's web api call Previously we had all of this running in PHP. Now we are experimenting with Rails, which is fantastic for #1 (the front end UI). The question is how to do #2, the back serving of widget information, efficiently. Obviously this is much higher load than the front end, since it is called every time the front page loads on one of our clients' websites. I can see two obvious approaches: A. Parallel Stack: Set up a parallel stack that uses something other than rails (e.g. our old PHP-based approach) but accesses the same database as the front end B. Rails Metal: Use Rails Metal/Rack to bypass the Rails routing mechanism, but keep the api call responder within the Rails app My main question: Is Rails/Metal a reasonable approach for something like this? But also... Will the overhead of loading the Rails environment still be too heavy? Is there a way to get even closer to the metal with Rails, bypassing most of the environment? Will Rails/Metal performance approach the perf of a similar task on straight PHP (just looking for ballpark here)? And... Is there a 'C' option that would be much better than both A and B? That is, something before going to the lengths of C code compiled to binary and installed as an nginx or apache module? Thanks in advance for any insights.

    Read the article

  • Data Transfer Objects VS Domain/ActiveRecord Entities in the View in RoR

    - by leypascua
    I'm coming from a .NET background, where it is a practice to not bind domain/entity models directly to the view in not-so-basic CRUD-ish applications where the view does not directly project entity fields as-is. I'm wondering what's the practice in RoR, where the default persistence mechanism is ActiveRecord. I would assert that presentation-related info should not be leaked to the entities, not sure though if this is how real RoR heads would do it. If DTOs/model per view is the approach, how will you do it in Rails? Your thoughts? EDIT: Some examples: - A view shows a list of invoices, with the number of unique items in one column. - A list of credit card accounts, where possibly fraudulent transactions were executed. For that, the UI needs to show this row in red. For both scenarios, The lists don't show all of the fields of the entities, just a few to show in the list (like invoice #, transaction date, name of the account, the amount of the transaction) For the invoice example, The invoice entity doesn't have a field "No. of line items" mapped on it. The database has not been denormalized for perf reasons and it will be computed during query time using aggregate functions. For the credit card accounts example, surely the card transaction entity doesn't have a "Show-in-red" or "IsFraudulent" invariant. Yes it may be a business rule, but for this example, that is a presentation concern, so I would like to keep it out of my domain model.

    Read the article

  • DNS-Based Environment Determination

    - by zvolkov
    Found the following here. The questions is: where can I find more details on how exactly implement this on Windows? Any guide or how-to anybody? Or maybe you can provide your invaluable suggestions? Specifically, how do I make so that "all QA servers would first resolve entries in qa.example.com first and then if that lookup failed they would try example.com" (I'm a dev, not a DNS specialist, but our IT Support has refused to help on this:() Use DNS Based Environment Determination for your servers. Do this by initially splitting your top level domain into a number of sub domains depending on their function, and then creating DNS Service Names in each of the sub domains pointing to the relevant server for that service. Based on the list above we would then have: * clientdb.prod.example.com for Production * clientdb.perf.example.com for Performance Testing * clientdb.qa.example.com for QA * clientdb.dev.example.com for Development Servers then resolve entries in their relevant sub domain by function. That is, all QA servers would first resolve entries in qa.example.com first and then if that lookup failed they would try example.com. This allows you to have a single configuration entry for your client database hostname (clientdb) that would resolve correctly in all environments. This technique has the added advantage of still having global services defined in a common top level domain. Here's one related (but not equivalent) SO question: http://stackoverflow.com/questions/774490/dns-resolving-based-on-client-ip This seems to be related to Providing "split horizon" DNS service. Reading that, I see that I will probably need separate DNS Server for each environment. Is this true or does Windows support some form of "tagging" the records to be visible depending on the requestor's IP? Also, cross-posted on ServerFault

    Read the article

  • x86 .net application with system.OutOfMemoryException

    - by Allen
    Hi,Guys I got OutOfMemoryException after the app running for 1 day, the app totally use 1.5G memory , all consumed by managed heap, gen 2 used 200mb , and LOB used 1.3mb, however the weired thing is, 900mb of space are Free. from perf counter I saw there had number of gen 2 gc collection happened, why GC collector cannot collect those 900mb free space in gen2 and LOB? I'm really appreicate for your help. following info are from windbg: 0:000> !eeheap -gc Number of GC Heaps: 1 generation 0 starts at 0x183153f0 generation 1 starts at 0x182aa834 generation 2 starts at 0x02131000 ephemeral segment allocation context: none segment begin allocated size 02130000 02131000 0312f284 0xffe284(16769668) 07750000 07751000 0874fc5c 0xffec5c(16772188) 09e30000 09e31000 0ae2fc2c 0xffec2c(16772140) 0b230000 0b231000 0c22ffec 0xffefec(16773100) 0c230000 0c231000 0d22f6f0 0xffe6f0(16770800) 0d230000 0d231000 0e22ea10 0xffda10(16767504) 0e230000 0e231000 0f22c1c4 0xffb1c4(16757188) 10390000 10391000 1138ddf4 0xffcdf4(16764404) 154e0000 154e1000 164da90c 0xff990c(16750860) 34aa0000 34aa1000 35a9dbfc 0xffcbfc(16763900) 7aca0000 7aca1000 7bc9edfc 0xffddfc(16768508) 49760000 49761000 4a75ef64 0xffdf64(16768868) 7bca0000 7bca1000 7cc99bac 0xff8bac(16747436) 17a70000 17a71000 183313fc 0x8c03fc(9176060) Large object heap starts at 0x03131000 segment begin allocated size 03130000 03131000 041250c8 0xff40c8(16728264) 08920000 08921000 099102f8 0xfef2f8(16708344) .... .... 4c760000 4c761000 4d71d578 0xfbc578(16500088) 1bb10000 1bb11000 1ca110d0 0xf000d0(15728848) 57760000 57761000 5862d7f8 0xecc7f8(15517688) Total Size: Size: 0x5ab13450 (1521562704) bytes. ------------------------------ GC Heap Size: Size: 0x5ab13450 (1521562704) bytes. 0:000> !dumpheap -stat total 0 objects Statistics: MT Count TotalSize Class Name 73037c78 1 12 System.Configuration.GenericEnumConverter 73036da0 1 12 System.Configuration.InfiniteIntConverter .... .... 69161c3c 35025 6809420 System.Windows.EffectiveValueEntry[] 69164748 54 12471072 MS.Internal.WeakEventTable+EventKey[] 710e2228 9540 190389260 System.Byte[] 710dd2b8 1317031 339257932 System.String 0035a670 6427 902224056 Free Total 3615631 objects

    Read the article

  • lots of backbone views - performance issues?

    - by ksol
    tl;dr: I wonder if having lots (100+ for the moment, potentially up to 1000/2000 or more) of backbone views (as a cell of a table) is too heavy or not The project I'm working on revolves around a planning view. There one row per user that covers 6 hours of a day, each hour splitted in 4 slots of 15mn. This planning is used to add "reservations" when clicking on a slot, and should handle hovering of the correct slots, and also handle when it is NOT possible to make a reservation - ie. prevent user click on an "unavailable" slot. There is many reasons why a slot can't be clicked on: the user is not available at this time, or the user is in a reservation; or the app needs to "force" a delay slot between two reservations. Reservations (a div) are rendered in a slot (a cell of a table), and by toying with dimensions, hovers the right number of slots. All this screen is handled with backbone. So For each slot I'm hovering on, I need to check wether I can do a reservation here or not. As of this moment, I use this by toying with the data attributes on the slots : when a reservation object is added, the slots covered are "enhanced with (among others) the reservation object (the backbone view object). But in some cases I don't quite have a grasp on now, it mixes up, and when the reservation view is removed, the slots are not "cleaned up" : the previous class is not reset correctly. It is probably something I've done wrong or badly, but this is only going to get heavier; I think I should use another class of Backbone views here, but I'm afraid the number of slots and thereof of views objects will be high and cause performance issue. I don't know mush about js perf so I'd like to have some feedback before jumping on that train. Any other advice on how to do this would be quite welcomed too. Thanks for your time. If this is not clear enough, tell me, I'll try and rephrase it.

    Read the article

  • Best practices for (over)using Azure queues

    - by John
    Hi, I'm in the early phases of designing an Azure-based application. One of the things that attracts me to Azure is the scalability, given the variability of the demand I'm likely to expect. As such I'm trying to keep things loosely coupled so I can add instances when I need to. The recommendations I've seen for architecting an application for Azure include keeping web role logic to a minimum, and having processing done in worker roles, using queues to communicate and some sort of back-end store like SQL Azure or Azure Tables. This seems like a good idea to me as I can scale up either or both parts of the application without any issue. However I'm curious if there are any best practices (or if anyone has any experiences) for when it's best to just have the web role talk directly to the data store vs. sending data by the queue? I'm thinking of the case where I have a simple insert to do from the web role - while I could set this up as a message, send it on the queue, and have a worker role pick it up and do the insert, it seems like a lot of double-handling. However I also appreciate that it may be the case that this is better in the long run, in case the web role gets overwhelmed or more complex logic ends up being required for the insert. I realise this might be a case where the answer is "it depends entirely on the situation, check your perf metrics" - but if anyone has any thoughts I'd be very appreciative! Thanks John

    Read the article

  • Best practice for debug Asserts during Unit testing

    - by Steve Steiner
    Does heavy use of unit tests discourage the use of debug asserts? It seems like a debug assert firing in the code under test implies the unit test shouldn't exist or the debug assert shouldn't exist. "There can be only one" seems like a reasonable principle. Is this the common practice? Or do you disable your debug asserts when unit testing, so they can be around for integration testing? Edit: I updated 'Assert' to debug assert to distinguish an assert in the code under test from the lines in the unit test that check state after the test has run. Also here is an example that I believe shows the dilema: A unit test passes invalid inputs for a protected function that asserts it's inputs are valid. Should the unit test not exist? It's not a public function. Perhaps checking the inputs would kill perf? Or should the assert not exist? The function is protected not private so it should be checking it's inputs for safety.

    Read the article

  • unbuffered I/O in Linux

    - by stuck
    I'm writing lots and lots of data that will not be read again for weeks - as my program runs the amount of free memory on the machine (displayed with 'free' or 'top') drops very quickly, the amount of memory my app uses does not increase - neither does the amount of memory used by other processes. This leads me to believe the memory is being consumed by the filesystems cache - since I do not intend to read this data for a long time I'm hoping to bypass the systems buffers, such that my data is written directly to disk. I dont have dreams of improving perf or being a super ninja, my hope is to give a hint to the filesystem that I'm not going to be coming back for this memory any time soon, so dont spend time optimizing for those cases. On Windows I've faced similar problems and fixed the problem using FILE_FLAG_NO_BUFFERING|FILE_FLAG_WRITE_THROUGH - the machines memory was not consumed by my app and the machine was more usable in general. I'm hoping to duplicate the improvements I've seen but on Linux. On Windows there is the restriction of writing in sector sized pieces, I'm happy with this restriction for the amount of gain I've measured. is there a similar way to do this in Linux?

    Read the article

  • 2010 Collaboration Summit Impressions

    - by Elena Zannoni
    It's a bit late, but there you have it anyway. April 14 to 16 I attended the Linux Foundation Collaboration Summit in SFO. I was running two tracks, one on tracing and one on tools. You can see the tracks and the slides here: http://events.linuxfoundation.org/events/collaboration-summit/slides I was pretty busy both days, Thursday with a whole day tracing track, Friday with a half day toolchain track. The sessions were well attended, the rooms were full, with people spilling in the hallways. Some new things were presented, like Kernelshark, by Steve Rostedt, a GUI (yes, believe it or not, a GUI) written in GTK. It is very nice, showing a timeline for traced kernel events, and you can zoom in and filter at will. It works on the latest kernels, and it requires some new things/fixes in GTK. I don't recall exactly what version of GTK though. Dominique Toupin from Ericsson presented something about user requirements for tracing. Mostly though about who's who in the embedded world, and eclipse. Masami and Mathieu presented an update on their work. See their slides. The interesting thing to me was of course the new version of uprobes w/o underlying utrace presented by Jim Keniston. At the end of the session we had a discussion about the future of utrace. Roland wasn't there, butTom Tromey (also from RedHat) collected the feedback. Basically we are at a standstill now that utrace has been rejected yet again. There wasn't much advise that anybody could give, except jokingly, we decided that the only way in is to make it a part of perf events. There needs to be another refactoring, but most of all, this "killer app" that would be enabled because of utrace hasn't materialized yet. We think that having a good debugging story on Linux is enough of a killer app, for instance allowing multiple tracers, and not relying on SIGCHLD etc. I think this wasn't completely clear to the kernel community. Trying to achieve debugging via a gdb stub inside the kernel interfacing to utrace and that is controlled via the gdb remote protocol also lost its appeal (thankfully, since the gdb remote protocol is archaic). Somebody would have to be creative in how to submit utrace. It doesn't have to be called utrace (it was really a random choice, for lack of a letter that was not already used in front of the word "trace"). So basically, I think the ideas behind utrace are sound, and the necessity of a new interface is acknowledged. But I believe the integration/submission process with the kernel folks has to restart from scratch, clean slate. We'll see. There are many conferences and meetings coming up in the near future where things can be discussed further. On the second day, Friday, we had the tools talks. It was interesting to observe the more "kernel" oriented people's behavior towards the gcc etc community. The first talk was by Mark Mitchell, about Gcc and its new plugin architecture. After that, Paolo talked about the new C++1x standard, which will be finalized in 2011. Many features are already implemented in the libstdc++ library and gcc and usable today. We had a few minutes (really, the half day track was quite short) where Bradley Kuhn from the Software Freedom Law Center explained the GPLv3 exception for gcc (due to the new gcc plugin architecture and the availability of the intermediate results from the compilation, which is a new thing). I will not try to explain, but basically you cannot take the result of the preprocessing and then use that in your own proprietary compiler. After, we had a talk by Ian Taylor about the new Gold linker. One good thing in that area is that they are trying to make gold the new default linker (for instance Fedora will use gold as the distro linker). However gold is very different from binutils' old linker. It doesn't use a linker script, for instance. The kernel has been linked with gold many times as an exercise (the ground work was done by Kris Van Hees), but this needs to be constantly tested/monitored because the kernel linker script is very complex, and uses esoteric features (Wenji is now monitoring that each kernel RC can be built with gold). It was positive that people are now aware of gold and the need for it to be ported to more architectures. It seems that the porting is very easy, with little arch dependent code. Finally Tom Tromey presented about gdb and the archer project. Archer is a development branch of gdb mostly done by RedHat, where they are focusing on better c++ printing, c++ expression parsing, and plugins. The archer work is merged regularly in the gdb mainline. In general it was a good conference. I did miss most of the first day, because that's when I flew in. But I caught a couple of talks. Nothing earth shattering, except for Google giving each person registered a free Android phone. Yey.

    Read the article

  • MVVM Light V4 preview 2 (BL0015) #mvvmlight

    - by Laurent Bugnion
    Over the past few weeks, I have worked hard on a few new features for MVVM Light V4. Here is a second early preview (consider this pre-alpha if you wish). The features are unit-tested, but I am now looking for feedback and there might be bugs! Bug correction: Messenger.CleanupList is now thread safe This was an annoying bug that is now corrected: In some circumstances, an exception could be thrown when the Messenger’s recipients list was cleaned up (i.e. the “dead” instances were removed). The method is called now and then and the exception was thrown apparently at random. In fact it was really a multi-threading issue, which is now corrected. Bug correction: AllowPartiallyTrustedCallers prevents EventToCommand to work This is a particularly annoying regression bug that was introduced in BL0014. In order to allow MVVM Light to work in XBAPs too, I added the AllowPartiallyTrustedCallers attribute to the assemblies. However, we just found out that this causes issues when using EventToCommand. In order to allow EventToCommand to continue working, I reverted to the previous state by removing the AllowPartiallyTrustedCallers attribute for now. I will work with my friends at Microsoft to try and find a solution. Stay tuned. Bug correction: XML documentation file is now generated in Release configuration The XML documentation file was not generated for the Release configuration. This was a simple flag in the project file that I had forgotten to set. This is corrected now. Applying EventToCommand to non-FrameworkElements This feature has been requested in order to be able to execute a command when a Storyboard is completed. I implemented this, but unfortunately found out that EventToCommand can only be added to Storyboards in Silverlight 3 and Silverlight 4, but not in WPF or in Windows Phone 7. This obviously limits the usefulness of this change, but I decided to publish it anyway, because it is pretty damn useful in Silverlight… Why not in WPF? In WPF, Storyboards added to a resource dictionary are frozen. This is a feature of WPF which allows to optimize certain objects for performance: By freezing them, it is a contract where we say “this object will not be modified anymore, so do your perf optimization on them without worrying too much”. Unfortunately, adding a Trigger (such as EventTrigger) to an object in resources does not work if this object is frozen… and unfortunately, there is no way to tell WPF not to freeze the Storyboard in the resources… so there is no way around that (at least none I can see. In Silverlight, objects are not frozen, so an EventTrigger can be added without problems. Why not in WP7? In Windows Phone 7, there is a totally different issue: Adding a Trigger can only be done to a FrameworkElement, which Storyboard is not. Here I think that we might see a change in a future version of the framework, so maybe this small trick will work in the future. Workaround? Since you cannot use the EventToCommand on a Storyboard in WPF and in WP7, the workaround is pretty obvious: Handle the Completed event in the code behind, and call the Command from there on the ViewModel. This object can be obtained by casting the DataContext to the ViewModel type. This means that the View needs to know about the ViewModel, but I never had issues with that anyway. New class: NotifyPropertyChanged Sometimes when you implement a model object (for example Customer), you would like to have it implement INotifyPropertyChanged, but without having all the frills of a ViewModelBase. A new class named NotifyPropertyChanged allows you to do that. This class is a simple implementation of INotifyPropertyChaned (with all the overloads of RaisePropertyChanged that were implemented in BL0014). In fact, ViewModelBase inherits NotifyPropertyChanged. ViewModelBase does not implement IDisposable anymore The IDisposable interface and the Dispose method had been marked obsolete in the ViewModelBase class already in V3. Now they have been removed. Note: By this, I do not mean that IDisposable is a bad interface, or that it shouldn’t be used on viewmodels. In the contrary, I know that this interface is very useful in certain circumstances. However, I think that having it by default on every instance of ViewModelBase was sending a wrong message. This interface has a strong meaning in .NET: After Dispose has been executed, the instance should not be used anymore, and should be ready for garbage collection. What I really wanted to have on ViewModelBase was rather a simple cleanup method, something that can be executed now and then during runtime. This is fulfilled by the ICleanup interface and its Cleanup method. If your ViewModels need IDisposable, you can still use it! You will just have to implement the interface on the class itself, because it is not available on ViewModelBase anymore. What’s next? I have a couple exciting new features implemented already but that need more testing before they go live… Just stay tuned and by MIX11 (12-14 April 2011), we should see at least a major addition to MVVM Light Toolkit, as well as another smaller feature which is pretty cool nonetheless More about this later! Happy Coding Laurent   Laurent Bugnion (GalaSoft) Subscribe | Twitter | Facebook | Flickr | LinkedIn

    Read the article

  • Do Not Optimize Without Measuring

    - by Alois Kraus
    Recently I had to do some performance work which included reading a lot of code. It is fascinating with what ideas people come up to solve a problem. Especially when there is no problem. When you look at other peoples code you will not be able to tell if it is well performing or not by reading it. You need to execute it with some sort of tracing or even better under a profiler. The first rule of the performance club is not to think and then to optimize but to measure, think and then optimize. The second rule is to do this do this in a loop to prevent slipping in bad things for too long into your code base. If you skip for some reason the measure step and optimize directly it is like changing the wave function in quantum mechanics. This has no observable effect in our world since it does represent only a probability distribution of all possible values. In quantum mechanics you need to let the wave function collapse to a single value. A collapsed wave function has therefore not many but one distinct value. This is what we physicists call a measurement. If you optimize your application without measuring it you are just changing the probability distribution of your potential performance values. Which performance your application actually has is still unknown. You only know that it will be within a specific range with a certain probability. As usual there are unlikely values within your distribution like a startup time of 20 minutes which should only happen once in 100 000 years. 100 000 years are a very short time when the first customer tries your heavily distributed networking application to run over a slow WIFI network… What is the point of this? Every programmer/architect has a mental performance model in his head. A model has always a set of explicit preconditions and a lot more implicit assumptions baked into it. When the model is good it will help you to think of good designs but it can also be the source of problems. In real world systems not all assumptions of your performance model (implicit or explicit) hold true any longer. The only way to connect your performance model and the real world is to measure it. In the WIFI example the model did assume a low latency high bandwidth LAN connection. If this assumption becomes wrong the system did have a drastic change in startup time. Lets look at a example. Lets assume we want to cache some expensive UI resource like fonts objects. For this undertaking we do create a Cache class with the UI themes we want to support. Since Fonts are expensive objects we do create it on demand the first time the theme is requested. A simple example of a Theme cache might look like this: using System; using System.Collections.Generic; using System.Drawing; struct Theme { public Color Color; public Font Font; } static class ThemeCache { static Dictionary<string, Theme> _Cache = new Dictionary<string, Theme> { {"Default", new Theme { Color = Color.AliceBlue }}, {"Theme12", new Theme { Color = Color.Aqua }}, }; public static Theme Get(string theme) { Theme cached = _Cache[theme]; if (cached.Font == null) { Console.WriteLine("Creating new font"); cached.Font = new Font("Arial", 8); } return cached; } } class Program { static void Main(string[] args) { Theme item = ThemeCache.Get("Theme12"); item = ThemeCache.Get("Theme12"); } } This cache does create font objects only once since on first retrieve of the Theme object the font is added to the Theme object. When we let the application run it should print “Creating new font” only once. Right? Wrong! The vigilant readers have spotted the issue already. The creator of this cache class wanted to get maximum performance. So he decided that the Theme object should be a value type (struct) to not put too much pressure on the garbage collector. The code Theme cached = _Cache[theme]; if (cached.Font == null) { Console.WriteLine("Creating new font"); cached.Font = new Font("Arial", 8); } does work with a copy of the value stored in the dictionary. This means we do mutate a copy of the Theme object and return it to our caller. But the original Theme object in the dictionary will have always null for the Font field! The solution is to change the declaration of struct Theme to class Theme or to update the theme object in the dictionary. Our cache as it is currently is actually a non caching cache. The funny thing was that I found out with a profiler by looking at which objects where finalized. I found way too many font objects to be finalized. After a bit debugging I found the allocation source for Font objects was this cache. Since this cache was there for years it means that the cache was never needed since I found no perf issue due to the creation of font objects. the cache was never profiled if it did bring any performance gain. to make the cache beneficial it needs to be accessed much more often. That was the story of the non caching cache. Next time I will write something something about measuring.

    Read the article

  • Which version of MSXML should I use?

    - by Cheeso
    Seems like this would be a common question, though I could not find it on SO. Which version of MSXML should I use in my applications, and more importantly, how should I decide? There is MSXML3, 4, 5 and 6. I recently posted some code in calling-wcf-service-by-vbscript that used MSXML v4. AnthonyWJones posted that I shouldn't use 4, but instead 3 or 6, but probably 3. Certainly not v5! Why? I'd like to know more about the criteria for selecting the version of MSXML to use in my apps. Bonus question: Does anyone have a summary of the differences between the various versions of MSXML over time? Summary so far: MSXML6 Should be first choice. was released in 2006, and includes perf and compliance fixes. Use this if you can. It's good. There are no merge modules; to bundle the MSXML6 runtime with your app, MS suggests packaging the MSXML6 msi file. MSXML6 is an upgrade from MSXML3/4 but does not replace them, because it discontinues some features. You can get the MSI here. MSXML3 Second choice. Most widely deployed version. Originally shipped in March 2000. Actively maintained, no new features. Currently supported, if you are on SP5 (shipped in 2005) or later. SP7 is current (also from 2005). MSXML5 was released only as part of MS-Office. Currently supported by Microsoft, but only as part of Office, not for building apps. Don't build apps that depend on MSXML5: Verboten. MSXML4 originally shipped? Currently in "maintenance mode". Microsoft is encouraging people to move off MSXML4 to MSXML6. Currently supported if you are on MSXML4SP2 or later, which shipped in 2003. download MSXML4SP2 here. Can be redisributed. Using the right version of MSXML in Internet Explorer is a good entry on the blog from Microsoft's xmlteam.

    Read the article

  • Should we denormalize database to improve performance?

    - by Groo
    We have a requirement to store 500 measurements per second, coming from several devices. Each measurement consists of a timestamp, a quantity type, and several vector values. Right now there is 8 vector values per measurement, and we may consider this number to be constant for needs of our prototype project. We are using HNibernate. Tests are done in SQLite (disk file db, not in-memory), but production will probably be MsSQL. Our Measurement entity class is the one that holds a single measurement, and looks like this: public class Measurement { public virtual Guid Id { get; private set; } public virtual Device Device { get; private set; } public virtual Timestamp Timestamp { get; private set; } public virtual IList<VectorValue> Vectors { get; private set; } } Vector values are stored in a separate table, so that each of them references its parent measurement through a foreign key. We have done a couple of things to ensure that generated SQL is (reasonably) efficient: we are using Guid.Comb for generating IDs, we are flushing around 500 items in a single transaction, ADO.Net batch size is set to 100 (I think SQLIte does not support batch updates? But it might be useful later). The problem Right now we can insert 150-200 measurements per second (which is not fast enough, although this is SQLite we are talking about). Looking at the generated SQL, we can see that in a single transaction we insert (as expected): 1 timestamp 1 measurement 8 vector values which means that we are actually doing 10x more single table inserts: 1500-2000 per second. If we placed everything (all 8 vector values and the timestamp) into the measurement table (adding 9 dedicated columns), it seems that we could increase our insert speed up to 10 times. Switching to SQL server will improve performance, but we would like to know if there might be a way to avoid unnecessary performance costs related to the way database is organized right now. [Edit] With in-memory SQLite I get around 350 items/sec (3500 single table inserts), which I believe is about as good as it gets with NHibernate (taking this post for reference: http://ayende.com/Blog/archive/2009/08/22/nhibernate-perf-tricks.aspx). But I might as well switch to SQL server and stop assuming things, right? I will update my post as soon as I test it.

    Read the article

  • generated service mock: everything but RhinoMocks fails?

    - by hko
    I have the "quest" to search for the next Mocking Framework for my company, and basically it's down to NSubstitute (simplest syntax, but no strict mocks), FakeItEasy(best reviews, Roy Osherove bonus, and slightly better lib support than NSubstitute), Moq (best "other libs support", biggest featureset, downside: mock.Object). We definitely want to move on from RhinoMocks, e.g. because of the unusefull interactiontest error messages (it should tell me what the parameter was instead, when a verification fails). So I was pretty surprised the other day (that was yesterday) when I found out RhinoMocks could do a thing where every other mock framework fails at: Mocking an autogenerated SomethingService (a typical VS autogenerated service with a default construtor in a partial class). Please don't argue about the design.. I intend to write lightweight integration tests (and some unit tests), and I can't mess around with the service, the product is installed on too many customers system. See this code: // here the NSubstitute and FakeItEasy equivalents throw an exception.. see below TicketStoreService fakeTicketStoreService = MockRepository.GenerateMock<TicketStoreService>(); fakeTicketStoreService.Expect(service => service.DoSomething(Arg.Is(new Guid())).Return(new Guid()); fakeTicketStoreService.DoSomething(Arg.Is(new Guid())); fakeTicketStoreService.VerifyAllExpectations(); Note that DoSomething is a non-virtual methodcall in an autogenerated class. So it shouldn't work, according to common knowledge. But it does. Problem is that it's the only (non commercial) framework that can do this: Rhino.Mocks works, and verification works too FakeItEasy says it doesn't find a default constructor (probably just wrong exception message): No default constructor was found on the type SomeNamespace.TicketStoreService Moq gives something sane and understandable: Invalid setup on a non-virtual (overridable in VB) member: service=> service.DoSomething Nsubstitute gives a message System.NotSupportedException: Cannot serialize member System.ComponentModel.Component.Site of type System.ComponentModel.ISite because it is an interface. I'm really wondering what's going on here with the frameworks, except Moq. The "fancy new" frameworks seem to have an initial perf hit too, probably preparing some Type cache and serializing stuff, whilst RhinoMocks somehow manages to create a very "slim" mock without recursion. I have to admit I didn't like RhinoMocks very well, but here it shines.. unfortunately. So, is there a way to get that to work with newer (non-commercial!) mocking frameworks, or somehow get a sane error message out of Rhino.Mocks? And why can Rhino.Mocks achieve this, when clearly every Mocking framework states it can only work with virtual methods when given a concrete class? Let's not derail the discussion by talking about alternative approaches like Extract&Override or runtime-proxy Mocking frameworks like JustMock/TypeMock/Moles or the new Fakes framework, I know these, but that would be less ideal solutions, for reasons beyond this topic. Any help appreciated..

    Read the article

< Previous Page | 1 2 3 4 5 6  | Next Page >