Search Results

Search found 15357 results on 615 pages for 'oracle corporation'.

Page 520/615 | < Previous Page | 516 517 518 519 520 521 522 523 524 525 526 527  | Next Page >

  • Take Two: Comparing JVMs on ARM/Linux

    - by user12608080
    Although the intent of the previous article, entitled Comparing JVMs on ARM/Linux, was to introduce and highlight the availability of the HotSpot server compiler (referred to as c2) for Java SE-Embedded ARM v7,  it seems, based on feedback, that everyone was more interested in the OpenJDK comparisons to Java SE-E.  In fact there were two main concerns: The fact that the previous article compared Java SE-E 7 against OpenJDK 6 might be construed as an unlevel playing field because version 7 is newer and therefore potentially more optimized. That the generic compiler settings chosen to build the OpenJDK implementations did not put those versions in a particularly favorable light. With those considerations in mind, we'll institute the following changes to this version of the benchmarking: In order to help alleviate an additional concern that there is some sort of benchmark bias, we'll use a different suite, called DaCapo.  Funded and supported by many prestigious organizations, DaCapo's aim is to benchmark real world applications.  Further information about DaCapo can be found at http://dacapobench.org. At the suggestion of Xerxes Ranby, who has been a great help through this entire exercise, a newer Linux distribution will be used to assure that the OpenJDK implementations were built with more optimal compiler settings.  The Linux distribution in this instance is Ubuntu 11.10 Oneiric Ocelot. Having experienced difficulties getting Ubuntu 11.10 to run on the original D2Plug ARMv7 platform, for these benchmarks, we'll switch to an embedded system that has a supported Ubuntu 11.10 release.  That platform is the Freescale i.MX53 Quick Start Board.  It has an ARMv7 Coretex-A8 processor running at 1GHz with 1GB RAM. We'll limit comparisons to 4 JVM implementations: Java SE-E 7 Update 2 c1 compiler (default) Java SE-E 6 Update 30 (c1 compiler is the only option) OpenJDK 6 IcedTea6 1.11pre 6b23~pre11-0ubuntu1.11.10.2 CACAO build 1.1.0pre2 OpenJDK 6 IcedTea6 1.11pre 6b23~pre11-0ubuntu1.11.10.2 JamVM build-1.6.0-devel Certain OpenJDK implementations were eliminated from this round of testing for the simple reason that their performance was not competitive.  The Java SE 7u2 c2 compiler was also removed because although quite respectable, it did not perform as well as the c1 compilers.  Recall that c2 works optimally in long-lived situations.  Many of these benchmarks completed in a relatively short period of time.  To get a feel for where c2 shines, take a look at the first chart in this blog. The first chart that follows includes performance of all benchmark runs on all platforms.  Later on we'll look more at individual tests.  In all runs, smaller means faster.  The DaCapo aficionado may notice that only 10 of the 14 DaCapo tests for this version were executed.  The reason for this is that these 10 tests represent the only ones successfully completed by all 4 JVMs.  Only the Java SE-E 6u30 could successfully run all of the tests.  Both OpenJDK instances not only failed to complete certain tests, but also experienced VM aborts too. One of the first observations that can be made between Java SE-E 6 and 7 is that, for all intents and purposes, they are on par with regards to performance.  While it is a fact that successive Java SE releases add additional optimizations, it is also true that Java SE 7 introduces additional complexity to the Java platform thus balancing out any potential performance gains at this point.  We are still early into Java SE 7.  We would expect further performance enhancements for Java SE-E 7 in future updates. In comparing Java SE-E to OpenJDK performance, among both OpenJDK VMs, Cacao results are respectable in 4 of the 10 tests.  The charts that follow show the individual results of those four tests.  Both Java SE-E versions do win every test and outperform Cacao in the range of 9% to 55%. For the remaining 6 tests, Java SE-E significantly outperforms Cacao in the range of 114% to 311% So it looks like OpenJDK results are mixed for this round of benchmarks.  In some cases, performance looks to have improved.  But in a majority of instances, OpenJDK still lags behind Java SE-Embedded considerably. Time to put on my asbestos suit.  Let the flames begin...

    Read the article

  • Look after your tribe of Pygmies with Java ME technology

    - by hinkmond
    Here's a game that is crossing over from the iDrone to the more lucrative Java ME cell phone market. See: Pocket God on Java ME Here's a quote: Massive casual iPhone hit Pocket God has parted the format waves and walked over to the land of Java mobiles, courtesy of AMA. The game sees you take control of an omnipotent, omnipresent, and (possibly) naughty deity, looking after your tribe of Pygmies... Everyone knows that there are more Java ME feature phones than grains of sand on a Pocket God island beach. So, when iDrone games are done piddlying around on a lesser platform, they move over to Java ME where things are really happening. Hinkmond

    Read the article

  • Selecting Items in a GeoToolkit Driven Map

    - by Geertjan
    When you take a look at all the tools provided by GeoToolkit, you'll be quite impressed. For example, within the US map shown in yesterday's blog entry, you can drill down into individual states by selecting them via the mouse, as shown below: With that, the basis of a more complex application is laid, since all the map-related functionality is handed to you out of the box. The sample referred to yesterday has been updated, if you check it out and run it (assuming you've taken the additional steps mentioned yesterday), you'll see the above. http://java.net/projects/nb-api-samples/sources/api-samples/show/versions/7.3/tutorials/geospatial/geotoolkit/MyGeospatialSystem

    Read the article

  • On-demand Webcast: Java in the Smart Grid

    - by Jacob Lehrbaum
    The Smart Grid is one of the most significant evolutions of our utility infrastructure in recent history. This innovative grid will soon revolutionize how utilities manage and control the energy in our homes--helping utilities reduce energy usage during peak hours, improve overall energy efficiency, and lower your energy bills. If you'd like to learn more about the Smart Grid and the role that Java is poised to play in this important initiative you can check out our on-demand webcast. We'll show you how Java solutions--including Java ME and Java SE for Embedded --can help build devices and infrastructure that take advantage of this new market. As the world's most popular developer language, Java enables you to work with a wide range of developers and provides access to tools and resources to build smarter devices, faster and more affordably.

    Read the article

  • RPi and Java Embedded GPIO: Connecting LEDs

    - by hinkmond
    Next, we need some low-level peripherals to connect to the Raspberry Pi GPIO header. So, we'll do what's called a "Fry's Run" in Silicon Valley, which means we go shop at the local Fry's Electronics store for parts. In this case, we'll need some breadboard jumper wires (blue wires in photo), some LEDs, and some resistors (for the RPi GPIO, 150 ohms - 300 ohms would work for the 3.3V output of the GPIO ports). And, if you want to do other projects, you might as well by a breadboard, which is a development board with lots of holes in it. Ask a Fry's clerk for help. Or, better yet, ask the customer standing next to you in the electronics components aisle for help. (Might be faster) So, go to your local hobby electronics store, or go to Fry's if you have one close by, and come back here to the next blog post to see how to hook these parts up. Hinkmond

    Read the article

  • Finding which activities will execute next in a process instance

    - by Mark Nelson
      We have had a few queries lately about how to find out what activity (or activities) will be the next to execute in a particular process instance.  It is possible to do this, however you will need to use a couple of undocumented APIs.  That means that they could (and probably will) change in some future release and break your code.  If you understand the risks of using undocumented APIs and are prepared to accept that risk, read on… READ MORE >>

    Read the article

  • JavaFX Makeover for JFugue Music NotePad

    - by Geertjan
    Bengt-Erik Fröberg from Sweden, one of the developers working on ProSang, the leading Scandinavian blood bank system (and based on the NetBeans Platform), is reworking the user interface of the JFugue Music NotePad. In particular, the Score window (named ScoreFX window below) contains components that are now quite clearly JavaFX, instead of Swing. Looks a lot better and also performs better. The sliders in the Keyboard window are candidates for being similarly redone to use JavaFX instead of Swing. Want to do something similar? Here's all the info you need: http://platform.netbeans.org/tutorials/nbm-javafx.html

    Read the article

  • Un-used Indexes on MDP_MATRIX Consuming Resources

    - by user702295
    Disable un-used Indexes: As much as it is recommended to create relevant indexes, it is advised not to have too many indexes on the mdp_matrix table.  Too many indexes will cause long waits on the table as indexes needs to get updated every time the table is updated.  There are many seeded indexes on mdp_matrix, every out of the box data model level has an index on the matrix table.  If a level is unused in the specific data model of the implementation, it is advisable to disable that index.  If the customer is not sure if and how indexes are utilized, the DBA can monitor all indexes.  After a few cycles of operation, the DBA should go over that list and see which indexes have not been used.  Consider disabling them. There are scripts on the net to monitor indexes or use the monitoring usage clause in the alter index statement.

    Read the article

  • Calgary SEO For Emerging Entrepreneurs

    With the economy as it is and job security becoming a rapidly fading memory, more and more people are starting to realize that they have just as much chance of success working for themselves as they do working for a large corporation. This is a great sign for the economy as an economy is always the most stable when people are being creative, innovative, and creating value on their own.

    Read the article

  • SQL Saturday 43 (Redmond, WA) Review

    - by BuckWoody
    Last Saturday (June 12th) we held a “SQL Saturday” (more about those here) event in Redmond, Washington. The event was held at the Microsoft campus, at the Mixer in our new location called the “Commons”. This is a mall-like area that we have on campus, and the Mixer is a large building with lots of meeting rooms, so it made a perfect location for the event. There was a sign to find the parking, and once there they had a sign to show how to get to the building. Since it’s a secure facility, Greg Larsen and crew had a person manning the door so that even late arrivals could get in. We had about 400 sign up for the event, and a little over 300 attend (official numbers later). I think we would have had a lot more, but the sun was out – and you just can’t underestimate the effect of that here in the Pacific Northwest. We joke a lot about not seeing the sun much, but when a day like what we had on Saturday comes around, and on a weekend at that, you’d cancel your wedding to go outside to play in the sun. And your spouse would agree with you for doing it. We had some top-notch speakers, including Clifford Dibble and Kalen Delany. The food was great, we had multiple sponsors (including Confio who seems to be at all of these) and the attendees were from all over the professional spectrum, from developers to BI to DBA’s. Everyone I saw was very engaged, and when I visited room-to-room I saw almost no one in the halls – everyone was in the sessions. I also saw a much larger Microsoft presence this year, especially from Dan Jones’ team. I had a great turnout at my session, and yes, I was wearing an Oracle staff shirt. I did that because I wanted to show that the session I gave on “SQL Server for the Oracle DBA” was non-marketing – I couldn’t exactly bash Oracle wearing their colors! These events are amazing. I can’t emphasize enough how much I appreciate the volunteers and how much work they put into these events, and to you for coming. If you’re reading this and you haven’t attended one yet, definitely find out if there is one in your area – and if not, start one. It’s a lot of work, but it’s totally worth it.       Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • EPM 11.1.2.1 - Smartview client and HFM office provider

    - by user809526
    If your connection to the smartview provider is very slow, because the login part takes a long time (user directory slowness, ...), consider adding on the desktop side a Windows parameter: HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\InternetSettings\ ReceiveTimeout 300000 to avoid being prompted over and over again for username/password This is an addition to the support doc id: "Smart View 11.1.2.1 Keeps Prompting For Username And Password For Financial Management Provider [ID 1353294.1]"

    Read the article

  • ZFS Basics

    - by user12614620
    Stage 1 basics: creating a pool # zpool create $NAME $REDUNDANCY $DISK1_0..N [$REDUNDANCY $DISK2_0..N]... $NAME = name of the pool you're creating. This will also be the name of the first filesystem and, by default, be placed at the mountpoint "/$NAME" $REDUNDANCY = either mirror or raidzN, and N can be 1, 2, or 3. If you leave N off, then it defaults to 1. $DISK1_0..N = the disks assigned to the pool. Example 1: zpool create tank mirror c4t1d0 c4t2d0 name of pool: tank redundancy: mirroring disks being mirrored: c4t1d0 and c4t2d0 Capacity: size of a single disk Example 2: zpool create tank raidz c4t1d0 c4t2d0 c4t3d0 c4t4d0 c4t5d0 Here the redundancy is raidz, and there are five disks, in a 4+1 (4 data, 1 parity) config. This means that the capacity is 4 times the disk size. If the command used "raidz2" instead, then the config would be 3+2. Likewise, "raidz3" would be a 2+3 config. Example 3: zpool create tank mirror c4t1d0 c4t2d0 mirror c4t3d0 c4t4d0 This is the same as the first mirror example, except there are two mirrors now. ZFS will stripe data across both mirrors, which means that writing data will go a bit faster. Note: you cannot create a mirror of two raidzs. You can create a raidz of mirrors, but to do that requires trickery.

    Read the article

  • Several New Hints

    - by Ondrej Brejla
    Hi all! Today we would like to introduce you some of our new experimental hints for NetBeans 7.2. They are called: Unused Use Statement and Immutable Variables. Unused Use Statement This hint is quite simple. It highlights (underlines) your use statements, which are not used. Typical use case is after some refactoring, when you forgot to remove some obsolete use statements. This hint warns you on them and allows you to remove them easily. Just click on the hint bulb in the gutter and select Remove Unused Use Statement. And of course, it works in multiple use statements combined too. Immutable Variables The next one is the hint which checks too many assignments into a variable. And why? That's simple. Mostly you should use just one assignment into one variable. But sometimes you are lazy and you do something like: But it's quite wrong, because what you really do is: And that's exactly the case, when our new hint warns you, that Too many assignments (2) into variable $foo occured. Nothing more. Yes, we know that there are some cases, where could be more assignments and no warning should occur, e.g.: Because maybe one likes longer increment syntax more than the short one. So we tried to handle these cases to don't bother you if it's not a need. Note: We are almost sure that this hint doesn't cover all your use cases, because there are a lot of them. So if you find something strange, write it into our bugzilla so we can handle it better for you. Thanks for your patience! And the last thing is, that you can set the number of allowed assignments in Tools -> Options -> Editor -> Hints -> PHP: Immutable Variables. Note: This hint works just for a common variables, not for fields. We have an enhancement request for that and it should be implemented in next version of NetBeans (probably 7.3). And that's all for today and as usual, please test it and if you find something strange, don't hesitate to file a new issue (product php, component Editor). Thanks.

    Read the article

  • Library order is important

    - by Darryl Gove
    I've written quite extensively about link ordering issues, but I've not discussed the interaction between archive libraries and shared libraries. So let's take a simple program that calls a maths library function: #include <math.h int main() { for (int i=0; i<10000000; i++) { sin(i); } } We compile and run it to get the following performance: bash-3.2$ cc -g -O fp.c -lm bash-3.2$ timex ./a.out real 6.06 user 6.04 sys 0.01 Now most people will have heard of the optimised maths library which is added by the flag -xlibmopt. This contains optimised versions of key mathematical functions, in this instance, using the library doubles performance: bash-3.2$ cc -g -O -xlibmopt fp.c -lm bash-3.2$ timex ./a.out real 2.70 user 2.69 sys 0.00 The optimised maths library is provided as an archive library (libmopt.a), and the driver adds it to the link line just before the maths library - this causes the linker to pick the definitions provided by the static library in preference to those provided by libm. We can see the processing by asking the compiler to print out the link line: bash-3.2$ cc -### -g -O -xlibmopt fp.c -lm /usr/ccs/bin/ld ... fp.o -lmopt -lm -o a.out... The flag to the linker is -lmopt, and this is placed before the -lm flag. So what happens when the -lm flag is in the wrong place on the command line: bash-3.2$ cc -g -O -xlibmopt -lm fp.c bash-3.2$ timex ./a.out real 6.02 user 6.01 sys 0.01 If the -lm flag is before the source file (or object file for that matter), we get the slower performance from the system maths library. Why's that? If we look at the link line we can see the following ordering: /usr/ccs/bin/ld ... -lmopt -lm fp.o -o a.out So the optimised maths library is still placed before the system maths library, but the object file is placed afterwards. This would be ok if the optimised maths library were a shared library, but it is not - instead it's an archive library, and archive library processing is different - as described in the linker and library guide: "The link-editor searches an archive only to resolve undefined or tentative external references that have previously been encountered." An archive library can only be used resolve symbols that are outstanding at that point in the link processing. When fp.o is placed before the libmopt.a archive library, then the linker has an unresolved symbol defined in fp.o, and it will search the archive library to resolve that symbol. If the archive library is placed before fp.o then there are no unresolved symbols at that point, and so the linker doesn't need to use the archive library. This is why libmopt needs to be placed after the object files on the link line. On the other hand if the linker has observed any shared libraries, then at any point these are checked for any unresolved symbols. The consequence of this is that once the linker "sees" libm it will resolve any symbols it can to that library, and it will not check the archive library to resolve them. This is why libmopt needs to be placed before libm on the link line. This leads to the following order for placing files on the link line: Object files Archive libraries Shared libraries If you use this order, then things will consistently get resolved to the archive libraries rather than to the shared libaries.

    Read the article

  • Lookup Viewer

    - by Geertjan
    The Maven integrated view that I showed yesterday I was able to create because I happened to know that an implementation of SubprojectProvider and LogicalViewProvider are in the Lookup of Maven projects. With that knowledge, I was able to use and even delegate to those implementations. But what if you don't know that those implementations are in the Lookup of the Project object? In the case of the Maven Project implementation, you could look in the source code of the Maven Project implementation, at the "getLookup" method. However, any other module could be putting its own objects into that Lookup, dynamically, i.e., at runtime. So there's no way of knowing what's in the Lookup of any Project object or any other object with a Lookup. But now imagine that you have a Lookup Viewer, as a tool during development, which you would exclude when distributing the application. Whenever new objects are found in the Lookup, the viewer displays them. You could install the Lookup Viewer into NetBeans IDE, or any other NetBeans Platform application, and then get a quick impression of what's actually in the Lookup when you select a different item in the application during development. Here it is (though I vaguely remember someone else writing something similar): Above, a Maven Project is selected. The Lookup Window shows that, among many other classes, an implementation of SubprojectProvider and LogicalViewProvider are found in the Lookup when the Maven Project is selected. If an item in the Lookup Window has its own Lookup, the content of that Lookup is displayed as child nodes of the Lookup, etc, i.e., you can explore all the way down the Lookup of each item found within objects found within the current selection. (What's especially fun is seeing the SaveCookieImpl being added and removed from the Lookup Window when you make/save a change in a document.) Another example is below, showing the Lookup Window installed in a custom application created during a course at MIT in Boston: A small trick I had to apply is that I always show the previous Lookup, since the current Lookup, when you select one of the Nodes in the Lookup Window, would be the Lookup of the Lookup Window itself! If anyone is interested in this, I can publish the NetBeans module providing the above window to the NetBeans update center. 

    Read the article

  • UPK version 3.1 goes GA!

    Hear Russell Handley, Director, UPK Product Marketing, discuss the much anticipated release of UPK 3.1, and how it can benefit enterprises of all sizes, across all geographies.

    Read the article

  • Community Forum at Openworld - Presentations available

    - by Javier Puerta
    On October 1st we held a new session of the Exadata & Manageability Partner Community in San Francisco. Thanks to all of you who participated in the event and very especially to the partner speakers who share their experiences with the rest of the community: Francisco Bermúdez (Capgemini Spain), Dmitry Krasilov (Nvision, Russia) and Miguel Alves (WeDo Technologies, Portugal)The slide decks used in the presentations are now available for download at the Manageability Partner Community Collaborative Workspace (for community members only - if you get an error message, please register for the Community first).In a few weeks we will be announcing the location for the next Community event in the spring timeframe.

    Read the article

  • Have You Checked Our BI Publisher Channel at Youtube ?

    - by kanichiro.nishida
    These days, more and more people watching video online rather than reading. Steve Jobs once said people don’t read anymore. Well, I love books and still read a lot either on books, magazine, iPad, MacbookPro, or whatever the medium shows me letters! But I have to admit, sometimes it’s much easier to understand especially something like How-To by just watching video clips than reading it. And this is why we started our BI Publisher Channel at Youtube last summer. Since then we have uploaded over 10 video clips so far and and now we’re gearing up to add more and more clips. Now, we’re in a middle of finishing up our work for the next 11G 1st patchset release, which should be coming soon and will have a lot of great new features that I can’t wait to talk to you guys about. And of course we’re preparing introduction and How-Top clips. So please subscribe the BI Publisher channel now if you haven’t done yet and stay tuned for the new clips! http://www.youtube.com/user/bipublisher Also, we’d love to hear your comments for each clip, so please don’t forget leaving your comments there after you watch!

    Read the article

  • Internet of Things (IoT) Thanksgiving Special: Turkey Tweeter (Part 1)

    - by hinkmond
    It's time for the Internet of Things (ioT) Thanksgiving Special. This time we are going to work on a special Do-It-Yourself project to create an Internet of Things temperature probe to connect your Turkey Day turkey to the Internet by writing a Thanksgiving Day Java Embedded app for your Raspberry Pi which will send out tweets as it cooks in your oven. If you're vegetarian, don't worry, you can follow along and just run the simulation of the Turkey Tweeter, or better yet, try a tofu version of the Turkey Tweeter. Here is the parts list: 1 Vernier Go!Temp USB Temperature Probe 1 Uncooked Turkey 1 Raspberry Pi (not Pumpkin Pie) 1 Roll thermal reflective tape You can buy the Vernier Go!Temp USB Temperature Probe for $39 from here: http://www.vernier.com/products/sensors/temperature-sensors/go-temp/. And, you can get the thermal reflective tape from any auto parts store. (Don't tell them what you need it for. Say it's for rebuilding your V-8 engine in your Dodge Hemi. Avoids the need for a long explanation and sounds cooler...) The uncooked turkey can be found in your neighborhood grocery store. But, if you're making a vegetarian Tofurkey, you're on your own... The Java Embedded app will be the same, though (Java is vegan). So, grab all your parts and come back here for the next part of this project... Hinkmond

    Read the article

  • Why are embedded device apps still written in C/C++? Why not Java programming language?

    - by hinkmond
    At the recent Black Hat 2014 conference in Sin City, the Black Hatters were focusing on Embedded Devices and IoT. You know? Make your networked-toaster burn your bread 10,000 miles away, over the Web for grins and giggles. Well, apparently the Black Hatters say it can be done pretty easily these days, which is scary. See: Securing Embedded Devices & IoT Here's a quote: All these devices are still written in C and C++. The challenges associated with developing securely in these languages have been fought for nearly two decades. "You often hear people say, 'Well, why don't we just get rid of the C and C++ language if it's so problematic. Why don't we just write everything in C# or Java, or something that is a little safer to develop in?'," DeMott says. Gah! Why are all these IoT devices still using C/C++? Of course they should be using Java SE Embedded technology! It's a natural fit to use for better security on embedded devices. Or, I guess, developers really don't mind if their networked-toasters do char their breakfast. If it can be burned, it will be... That's what I say. Unless they use Java. Hinkmond

    Read the article

  • Smarty: Tags Matching and Unpaired Tags Errors

    - by Martin Fousek
    Hello, today we would like to show you other improvements we have prepared in PHP Smarty Framework. Let's talk about highlighting of matching tags and error reporting of unpaired ones. Tags Matching Some of your enhancements talked  about paired tags matching to be able to see matching tags at first glance.We have good news for you that this feature you can try out already in our latest PHP Development builds and of course later in NetBeans 7.3. Unpaired Tags Errors To make easier detecting of template syntax issues, we provide basic tags pairing. If you forgot to begin some paired Smarty tag or you end it unexpectedly you should get error hint which complains about your issue. That's all for today. As always, please test it and report all the issues or enhancements you find in NetBeans BugZilla (component php, subcomponent Smarty).

    Read the article

  • ZFS for Database Log Files

    - by user12620111
    I've been troubled by drop outs in CPU usage in my application server, characterized by the CPUs suddenly going from close to 90% CPU busy to almost completely CPU idle for a few seconds. Here is an example of a drop out as shown by a snippet of vmstat data taken while the application server is under a heavy workload. # vmstat 1  kthr      memory            page            disk          faults      cpu  r b w   swap  free  re  mf pi po fr de sr s3 s4 s5 s6   in   sy   cs us sy id  1 0 0 130160176 116381952 0 16 0 0 0 0  0  0  0  0  0 207377 117715 203884 70 21 9  12 0 0 130160160 116381936 0 25 0 0 0 0 0  0  0  0  0 200413 117162 197250 70 20 9  11 0 0 130160176 116381920 0 16 0 0 0 0 0  0  1  0  0 203150 119365 200249 72 21 7  8 0 0 130160176 116377808 0 19 0 0 0 0  0  0  0  0  0 169826 96144 165194 56 17 27  0 0 0 130160176 116377800 0 16 0 0 0 0  0  0  0  0  1 10245 9376 9164 2  1 97  0 0 0 130160176 116377792 0 16 0 0 0 0  0  0  0  0  2 15742 12401 14784 4 1 95  0 0 0 130160176 116377776 2 16 0 0 0 0  0  0  1  0  0 19972 17703 19612 6 2 92  14 0 0 130160176 116377696 0 16 0 0 0 0 0  0  0  0  0 202794 116793 199807 71 21 8  9 0 0 130160160 116373584 0 30 0 0 0 0  0  0 18  0  0 203123 117857 198825 69 20 11 This behavior occurred consistently while the application server was processing synthetic transactions: HTTP requests from JMeter running on an external machine. I explored many theories trying to explain the drop outs, including: Unexpected JMeter behavior Network contention Java Garbage Collection Application Server thread pool problems Connection pool problems Database transaction processing Database I/O contention Graphing the CPU %idle led to a breakthrough: Several of the drop outs were 30 seconds apart. With that insight, I went digging through the data again and looking for other outliers that were 30 seconds apart. In the database server statistics, I found spikes in the iostat "asvc_t" (average response time of disk transactions, in milliseconds) for the disk drive that was being used for the database log files. Here is an example:                     extended device statistics     r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 2053.6    0.0 8234.3  0.0  0.2    0.0    0.1   0  24 c3t60080E5...F4F6d0s0     0.0 2162.2    0.0 8652.8  0.0  0.3    0.0    0.1   0  28 c3t60080E5...F4F6d0s0     0.0 1102.5    0.0 10012.8  0.0  4.5    0.0    4.1   0  69 c3t60080E5...F4F6d0s0     0.0   74.0    0.0 7920.6  0.0 10.0    0.0  135.1   0 100 c3t60080E5...F4F6d0s0     0.0  568.7    0.0 6674.0  0.0  6.4    0.0   11.2   0  90 c3t60080E5...F4F6d0s0     0.0 1358.0    0.0 5456.0  0.0  0.6    0.0    0.4   0  55 c3t60080E5...F4F6d0s0     0.0 1314.3    0.0 5285.2  0.0  0.7    0.0    0.5   0  70 c3t60080E5...F4F6d0s0 Here is a little more information about my database configuration: The database and application server were running on two different SPARC servers. Storage for the database was on a storage array connected via 8 gigabit Fibre Channel Data storage and log file were on different physical disk drives Reliable low latency I/O is provided by battery backed NVRAM Highly available: Two Fibre Channel links accessed via MPxIO Two Mirrored cache controllers The log file physical disks were mirrored in the storage device Database log files on a ZFS Filesystem with cutting-edge technologies, such as copy-on-write and end-to-end checksumming Why would I be getting service time spikes in my high-end storage? First, I wanted to verify that the database log disk service time spikes aligned with the application server CPU drop outs, and they did: At first, I guessed that the disk service time spikes might be related to flushing the write through cache on the storage device, but I was unable to validate that theory. After searching the WWW for a while, I decided to try using a separate log device: # zpool add ZFS-db-41 log c3t60080E500017D55C000015C150A9F8A7d0 The ZFS log device is configured in a similar manner as described above: two physical disks mirrored in the storage array. This change to the database storage configuration eliminated the application server CPU drop outs: Here is the zpool configuration: # zpool status ZFS-db-41   pool: ZFS-db-41  state: ONLINE  scan: none requested config:         NAME                                     STATE         ZFS-db-41                                ONLINE           c3t60080E5...F4F6d0  ONLINE         logs           c3t60080E5...F8A7d0  ONLINE Now, the I/O spikes look like this:                     extended device statistics                  r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 1053.5    0.0 4234.1  0.0  0.8    0.0    0.7   0  75 c3t60080E5...F8A7d0s0                     extended device statistics                  r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 1131.8    0.0 4555.3  0.0  0.8    0.0    0.7   0  76 c3t60080E5...F8A7d0s0                     extended device statistics                  r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 1167.6    0.0 4682.2  0.0  0.7    0.0    0.6   0  74 c3t60080E5...F8A7d0s0     0.0  162.2    0.0 19153.9  0.0  0.7    0.0    4.2   0  12 c3t60080E5...F4F6d0s0                     extended device statistics                  r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 1247.2    0.0 4992.6  0.0  0.7    0.0    0.6   0  71 c3t60080E5...F8A7d0s0     0.0   41.0    0.0   70.0  0.0  0.1    0.0    1.6   0   2 c3t60080E5...F4F6d0s0                     extended device statistics                  r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 1241.3    0.0 4989.3  0.0  0.8    0.0    0.6   0  75 c3t60080E5...F8A7d0s0                     extended device statistics                  r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device     0.0 1193.2    0.0 4772.9  0.0  0.7    0.0    0.6   0  71 c3t60080E5...F8A7d0s0 We can see the steady flow of 4k writes to the ZIL device from O_SYNC database log file writes. The spikes are from flushing the transaction group. Like almost all problems that I run into, once I thoroughly understand the problem, I find that other people have documented similar experiences. Thanks to all of you who have documented alternative approaches. Saved for another day: now that the problem is obvious, I should try "zfs:zfs_immediate_write_sz" as recommended in the ZFS Evil Tuning Guide. References: The ZFS Intent Log Solaris ZFS, Synchronous Writes and the ZIL Explained ZFS Evil Tuning Guide: Cache Flushes ZFS Evil Tuning Guide: Tuning ZFS for Database Performance

    Read the article

< Previous Page | 516 517 518 519 520 521 522 523 524 525 526 527  | Next Page >