Search Results

Search found 12398 results on 496 pages for 'in memory oltp'.

Page 50/496 | < Previous Page | 46 47 48 49 50 51 52 53 54 55 56 57  | Next Page >

  • What's the purpose of having a separate "operator new[]" ?

    - by sharptooth
    Looks like operator new and operator new[] have exactly the same signature: void* operator new( size_t size ); void* operator new[]( size_t size ); and do exactly the same: either return a pointer to a big enough block of raw (not initialized in any way) memory or throw an exception. Also operator new is called internally when I create an object with new and operator new[] - when I create an array of objects with new[]. Still the above two special functions are called by C++ internally in exactly the same manner and I don't se how the two calls can have different meanings. What's the purpose of having two different functions with exactly the same signatures and exactly the same behavior?

    Read the article

  • _CrtDumpMemoryLeaks truncated output??

    - by Marin
    Hello, I am trying to use Visual Studio's capability to detect memory leaks, but I keep getting truncated output, like: Dumping objects -> {174} normal block at 0x0099ADB8, 48 bytes long. Data: <h:\najnovije\tru> 68 3A 5C 6E 61 6A 6E 6F 76 69 6A 65 5C 74 72 75 {170} normal block at 0x0099AD58, 32 bytes long. Data: <h:\najnovije\tru> 68 3A 5C 6E 61 6A 6E 6F 76 69 6A 65 5C 74 72 75 Object dump complete. What am I doing wrong? I added #define _CRTDBG_MAP_ALLOC #include <stdlib.h> #include <crtdbg.h> to the beginning of my code. Thank you.

    Read the article

  • How can i see if dealloc is being called on a uikit object, or any object not created by myself

    - by Gyozo Kudor
    I think i have an UIImage that has a higher retain count than it should have and i am probably leaking memory. I use this image as a thumbnail, to set a custom background to a uibutton. So the uibutton is holding a reference to it and so do i. But instead of 2, the retainCount is 3. Do i have to create a custom UIImage derived class and override dealloc if I want to place a log message there and then change the class used from UIImage to my class, or is there an easier way. Thanks in advance.

    Read the article

  • Can I switch the Visual C++ runtime to another heap?

    - by sharptooth
    My program uses a third party dynamic link library that has huge memory leaks inside. Both my program and the library are Visual C++ native code. Both link to the Visual C++ runtime dynamically. I'd like to force the library into another heap so that all allocations that are done through the Visual C++ runtime while the library code is running are done on that heap. I can call HeapCreate() and later HeapDestroy(). If I somehow ensure that all allocations are done in the new heap I don't care of the leaks anymore - they all go when I destroy the second heap. Is it possible to force the Visual C++ runtime to make all allocations on a specified heap?

    Read the article

  • strategy to allocate/free lots of small objects

    - by aaa
    hello I am toying with certain caching algorithm, which is challenging somewhat. Basically, it needs to allocate lots of small objects (double arrays, < 256 elements), with objects accessible through mapped value, map[key] = array. time to initialized array may be quite large, generally more than 10 thousand cpu cycles. By lots I mean around gigabyte in total. objects may need to be popped/pushed as needed, generally in random places, one object at a time. lifetime of an object is generally long, minutes or more, however, object may be subject to allocation/deallocation several times during duration of program. What would be good strategy to avoid memory fragmentation, while still maintaining reasonable allocate deallocate speed? I am using C++, so I can use new and malloc. Thanks. I know there a similar questions on website, http://stackoverflow.com/questions/2156745/efficiently-allocating-many-short-lived-small-objects, are somewhat different, thread safety is not immediate issue for me.

    Read the article

  • Why does my motherboard go through an endless reboot cycle when 8 GB of memory is attempted vs 6 GB?

    - by nizm0
    I never got an answer in my googling to this about a year ago and have an extra stick of memory I'd like to be able to use. When is inserted the computer starts, and then reboots immediately in an endless reboot cycle. As soon as the 4th stick is removed, the computer works fine. Right now I have 6 GB of my 8 GB installed. Is there a solution that I am missing to enabling this motherboard to actually boot up with all 8 GB (it supports it). Right now it won't even boot up to BIOS with the 4 sticks... only 3? Memory: 1 x G.SKILL 4 GB (2 x 2 GB) 240-Pin DDR2 SDRAM DDR2 1100 (PC2 8800) Dual Channel Kit Desktop Memory Model F2-8800CL5D-4GBPI - Retail (URL: ) http://www.newegg.com/Product/Product.aspx?Item=N82E16820231194 Motherboard: 1 x GIGABYTE GA-EP45-UD3R LGA 775 Intel P45 ATX Intel Motherboard - Retail ( URL: )http://www.newegg.com/Product/Product.aspx?Item=N82E16813128359

    Read the article

  • Using the Static Code Analysis feature of Visual Studio (Premium/Ultimate) to find memory leakage problems

    - by terje
    Memory for managed code is handled by the garbage collector, but if you use any kind of unmanaged code, like native resources of any kind, open files, streams and window handles, your application may leak memory if these are not properly handled.  To handle such resources the classes that own these in your application should implement the IDisposable interface, and preferably implement it according to the pattern described for that interface. When you suspect a memory leak, the immediate impulse would be to start up a memory profiler and start digging into that.   However, before you follow that impulse, do a Static Code Analysis run with a ruleset tuned to finding possible memory leaks in your code.  If you get any warnings from this, fix them before you go on with the profiling. How to use a ruleset In Visual Studio 2010 (Premium and Ultimate editions) you can define your own rulesets containing a list of Static Code Analysis checks.   I have defined the memory checks as shown in the lists below as ruleset files, which can be downloaded – see bottom of this post.  When you get them, you can easily attach them to every project in your solution using the Solution Properties dialog. Right click the solution, and choose Properties at the bottom, or use the Analyze menu and choose “Configure Code Analysis for Solution”: In this dialog you can now choose the Memorycheck ruleset for every project you want to investigate.  Pressing Apply or Ok opens every project file and changes the projects code analysis ruleset to the one we have specified here. How to define your own ruleset  (skip this if you just download my predefined rulesets) If you want to define the ruleset yourself, open the properties on any project, choose Code Analysis tab near the bottom, choose any ruleset in the drop box and press Open Clear out all the rules by selecting “Source Rule Sets” in the Group By box, and unselect the box Change the Group By box to ID, and select the checks you want to include from the lists below. Note that you can change the action for each check to either warning, error or none, none being the same as unchecking the check.   Now go to the properties window and set a new name and description for your ruleset. Then save (File/Save as) the ruleset using the new name as its name, and use it for your projects as detailed above. It can also be wise to add the ruleset to your solution as a solution item. That way it’s there if you want to enable Code Analysis in some of your TFS builds.   Running the code analysis In Visual Studio 2010 you can either do your code analysis project by project using the context menu in the solution explorer and choose “Run Code Analysis”, you can define a new solution configuration, call it for example Debug (Code Analysis), in for each project here enable the Enable Code Analysis on Build   In Visual Studio Dev-11 it is all much simpler, just go to the Solution root in the Solution explorer, right click and choose “Run code analysis on solution”.     The ruleset checks The following list is the essential and critical memory checks.  CheckID Message Can be ignored ? Link to description with fix suggestions CA1001 Types that own disposable fields should be disposable No  http://msdn.microsoft.com/en-us/library/ms182172.aspx CA1049 Types that own native resources should be disposable Only if the pointers assumed to point to unmanaged resources point to something else  http://msdn.microsoft.com/en-us/library/ms182173.aspx CA1063 Implement IDisposable correctly No  http://msdn.microsoft.com/en-us/library/ms244737.aspx CA2000 Dispose objects before losing scope No  http://msdn.microsoft.com/en-us/library/ms182289.aspx CA2115 1 Call GC.KeepAlive when using native resources See description  http://msdn.microsoft.com/en-us/library/ms182300.aspx CA2213 Disposable fields should be disposed If you are not responsible for release, of if Dispose occurs at deeper level  http://msdn.microsoft.com/en-us/library/ms182328.aspx CA2215 Dispose methods should call base class dispose Only if call to base happens at deeper calling level  http://msdn.microsoft.com/en-us/library/ms182330.aspx CA2216 Disposable types should declare a finalizer Only if type does not implement IDisposable for the purpose of releasing unmanaged resources  http://msdn.microsoft.com/en-us/library/ms182329.aspx CA2220 Finalizers should call base class finalizers No  http://msdn.microsoft.com/en-us/library/ms182341.aspx Notes: 1) Does not result in memory leak, but may cause the application to crash   The list below is a set of optional checks that may be enabled for your ruleset, because the issues these points too often happen as a result of attempting to fix up the warnings from the first set.   ID Message Type of fault Can be ignored ? Link to description with fix suggestions CA1060 Move P/invokes to NativeMethods class Security No http://msdn.microsoft.com/en-us/library/ms182161.aspx CA1816 Call GC.SuppressFinalize correctly Performance Sometimes, see description http://msdn.microsoft.com/en-us/library/ms182269.aspx CA1821 Remove empty finalizers Performance No http://msdn.microsoft.com/en-us/library/bb264476.aspx CA2004 Remove calls to GC.KeepAlive Performance and maintainability Only if not technically correct to convert to SafeHandle http://msdn.microsoft.com/en-us/library/ms182293.aspx CA2006 Use SafeHandle to encapsulate native resources Security No http://msdn.microsoft.com/en-us/library/ms182294.aspx CA2202 Do not dispose of objects multiple times Exception (System.ObjectDisposedException) No http://msdn.microsoft.com/en-us/library/ms182334.aspx CA2205 Use managed equivalents of Win32 API Maintainability and complexity Only if the replace doesn’t provide needed functionality http://msdn.microsoft.com/en-us/library/ms182365.aspx CA2221 Finalizers should be protected Incorrect implementation, only possible in MSIL coding No http://msdn.microsoft.com/en-us/library/ms182340.aspx   Downloadable ruleset definitions I have defined three rulesets, one called Inmeta.Memorycheck with the rules in the first list above, and Inmeta.Memorycheck.Optionals containing the rules in the second list, and the last one called Inmeta.Memorycheck.All containing the sum of the two first ones.  All three rulesets can be found in the  zip archive  “Inmeta.Memorycheck” downloadable from here.   Links to some other resources relevant to Static Code Analysis MSDN Magazine Article by Mickey Gousset on Static Code Analysis in VS2010 MSDN :  Analyzing Managed Code Quality by Using Code Analysis, root of the documentation for this Preventing generated code from being analyzed using attributes Online training course on Using Code Analysis with VS2010 Blogpost by Tatham Oddie on custom code analysis rules How to write custom rules, from Microsoft Code Analysis Team Blog Microsoft Code Analysis Team Blog

    Read the article

  • Is it possible to have a computer run two OS's in memory at the same time?

    - by Hebon
    I'm sick of needing to reboot my computer every time I wish to use another OS, or run a virtual machine that skimps on power. With the onset of large amounts of memory for computers nowadays I began to think that there must be some way to run two OS's in memory with a way to switch between the two. In my mind, it doesn't seem too difficult; a compatibility layer boots up after bios, which in turn boots to OS1. While in OS1, software is run that triggers a save to ram boots back to the compatability layer, and then boots to OS2. This way, the OS's can be used side by side and boot times are cut drastically short since both OSs are already in ram. Both OS's have their own designated and protected memory so there is no problem there... I mean, it seems fine, but no one has done it, so there must be some reason as to why. I would love some insight into this please.

    Read the article

  • Why is the usable memory on my Macbook pro shown as 2.74 Gb when there is 4GB installed with 32bit Windows 7? [closed]

    - by Bobby Alexander
    Possible Duplicate: Windows XP and RAM 3.5GB+ Installed RAM : 4 GB but 2.96GB Usable......why? I have a Macbook Pro with 4GB of installed RAM. I have installed Windows 7 on it which shows the usable memory as 2.74GB. Why is this? Don't tell me the 32 bit story; I program for a living. The maximum addressible memory on a 32 bit system is 4 GB not 3 GB. Need proof? MSDN: Memory Limits for Windows Releases

    Read the article

  • Differences in memory consumption between two identical D7 sites?

    - by aendrew
    I'm running Drupal on a news site that has a lot of different View blocks on the front page (~5 total, all cached). In trying to reduce the memory footprint of the site, I've checked out source from SVN to a local development install to try and convert some of those blocks into more optimized code. Here's the weird thing. Devel module lists memory consumption at 50mb on the Production site (Running Nginx, PHP 5.2.17, XCache and Zend Optimizer.) but only 14mb on my development site (Running Apache2, PHP 5.2.13 and XCache). These are nearly-identical versions of the same site — frankly, the Production site should use even less memory as I've disabled some of the modules running on the Dev site. Any idea why this might be the case?

    Read the article

  • Memory Efficient Windows SOA Server

    - by Antony Reynolds
    Installing a Memory Efficient SOA Suite 11.1.1.6 on Windows Server Well 11.1.1.6 is now available for download so I thought I would build a Windows Server environment to run it.  I will minimize the memory footprint of the installation by putting all functionality into the Admin Server of the SOA Suite domain. Required Software 64-bit JDK SOA Suite If you want 64-bit then choose “Generic” rather than “Microsoft Windows 32bit JVM” or “Linux 32bit JVM” This has links to all the required software. If you choose “Generic” then the Repository Creation Utility link does not show, you still need this so change the platform to “Microsoft Windows 32bit JVM” or “Linux 32bit JVM” to get the software. Similarly if you need a database then you need to change the platform to get the link to XE for Windows or Linux. If possible I recommend installing a 64-bit JDK as this allows you to assign more memory to individual JVMs. Windows XE will work, but it is better if you can use a full Oracle database because of the limitations on XE that sometimes cause it to run out of space with large or multiple SOA deployments. Installation Steps The following flow chart outlines the steps required in installing and configuring SOA Suite. The steps in the diagram are explained below. 64-bit? Is a 64-bit installation required?  The Windows & Linux installers will install 32-bit versions of the Sun JDK and JRockit.  A separate JDK must be installed for 64-bit. Install 64-bit JDK The 64-bit JDK can be either Hotspot or JRockit.  You can choose either JDK 1.7 or 1.6. Install WebLogic If you are using 64-bit then install WebLogic using “java –jar wls1036_generic.jar”.  Make sure you include Coherence in the installation, the easiest way to do this is to accept the “Typical” installation. SOA Suite Required? If you are not installing SOA Suite then you can jump straight ahead and create a WebLogic domain. Install SOA Suite Run the SOA Suite installer and point it at the existing Middleware Home created for WebLogic.  Note to run the SOA installer on Windows the user must have admin privileges.  I also found that on Windows Server 2008R2 I had to start the installer from a command prompt with administrative privileges, granting it privileges when it ran caused it to ignore the jreLoc parameter. Database Available? Do you have access to a database into which you can install the SOA schema.  SOA Suite requires access to an Oracle database (it is supported on other databases but I would always use an oracle database). Install Database I use an 11gR2 Oracle database to avoid XE limitations.  Make sure that you set the database character set to be unicode (AL32UTF8).  I also disabled the new security settings because they get in the way for a developer database.  Don’t forget to check that number of processes is at least 150 and number of sessions is not set, or is set to at least 200 (in the DB init parameters). Run RCU The SOA Suite database schemas are created by running the Repository Creation Utility.  Install the “SOA and BPM Infrastructure” component to support SOA Suite.  If you keep the schema prefix as “DEV” then the config wizard is easier to complete. Run Config Wizard The Config wizard creates the domain which hosts the WebLogic server instances.  To get a minimum footprint SOA installation choose the “Oracle Enterprise Manager” and “Oracle SOA Suite for developers” products.  All other required products will be automatically selected. The “for developers” installs target the appropriate components at the AdminServer rather than creating a separate managed server to house them.  This reduces the number of JVMs required to run the system and hence the amount of memory required.  This is not suitable for anything other than a developer environment as it mixes the admin and runtime functions together in a single server.  It also takes a long time to load all the required modules, making start up a slow process. If it exists I would recommend running the config wizard found in the “oracle_common/common/bin” directory under the middleware home.  This should have access to all the templates, including SOA. If you also want to run BAM in the same JVM as everything else then you need to “Select Optional Configuration” for “Managed Servers, Clusters and Machines”. To target BAM at the AdminServer delete the “bam_server1” managed server that is created by default.  This will result in BAM being targeted at the AdminServer. Installation Issues I had a few problems when I came to test everything in my mega-JVM. Following applications were not targeted and so I needed to target them at the AdminServer: b2bui composer Healthcare UI FMW Welcome Page Application (11.1.0.0.0) How Memory Efficient is It? On a Windows 2008R2 Server running under VirtualBox I was able to bring up both the 11gR2 database and SOA/BPM/BAM in 3G memory.  I allocated a minimum 512M to the PermGen and a minimum of 1.5G for the heap.  The setting from setSOADomainEnv are shown below: set DEFAULT_MEM_ARGS=-Xms1536m -Xmx2048m set PORT_MEM_ARGS=-Xms1536m -Xmx2048m set DEFAULT_MEM_ARGS=%DEFAULT_MEM_ARGS% -XX:PermSize=512m -XX:MaxPermSize=768m set PORT_MEM_ARGS=%PORT_MEM_ARGS% -XX:PermSize=512m -XX:MaxPermSize=768m I arrived at these numbers by monitoring JVM memory usage in JConsole. Task Manager showed total system memory usage at 2.9G – just below the 3G I allocated to the VM. Performance is not stellar but it runs and I could run JDeveloper alongside it on my 8G laptop, so in that sense it was a result!

    Read the article

  • Strategies to Defeat Memory Editors for Cheating - Desktop Games

    - by ashes999
    I'm assuming we're talking about desktop games -- something the player downloads and runs on their local computer. Many are the memory editors that allow you to detect and freeze values, like your player's health. How do you prevent cheating via memory-modifiation? What strategies are effective to combat this kind of cheating? For reference, I know that players can: - Search for something by value or range - Search for something that changed value - Set memory values - Freeze memory values I'm looking for some good ones. Two I use that are mediocre are: Displaying values as a percentage instead of the number (eg. 46/50 = 92% health) A low-level class that holds values in an array and moves them with each change. (For example, instead of an int, I have a class that's an array of ints, and whenever the value changes, I use a different, randomly-chosen array item to hold the value)

    Read the article

  • Identify memory leak in Java app

    - by Vincent Ma
    One important advantage of java is programer don't care memory management and GC handle it well. Maybe this is one reason why java is more popular. As Java programer you real dont care it? After you meet Out of memory you will realize it it’s not true. Java GC and memory is big topic you can get some information in here Today just let me show how to identify memory leak quickly. Let quickly review demo java code, it’s one kind of memory leak in our code, using static collection and always add some object. import java.util.ArrayList;import java.util.List; public class MemoryTest { public static void main(String[] args) { new Thread(new MemoryLeak(), "MemoryLeak").start(); }} class MemoryLeak implements Runnable { public static List<Integer> leakList = new ArrayList<Integer>(); public void run() { int num =0; while(true) { try { Thread.sleep(1); } catch (InterruptedException e) { } num++; Integer i = new Integer(num); leakList.add(i); } }} run it with java -verbose:gc -XX:+PrintGCDetails -Xmx60m -XX:MaxPermSize=160m MemoryTest after about some minuts you will get Exception in thread "MemoryLeak" java.lang.OutOfMemoryError: Java heap space at java.util.Arrays.copyOf(Arrays.java:2760) at java.util.Arrays.copyOf(Arrays.java:2734) at java.util.ArrayList.ensureCapacity(ArrayList.java:167) at java.util.ArrayList.add(ArrayList.java:351) at MemoryLeak.run(MemoryTest.java:25) at java.lang.Thread.run(Thread.java:619)Heap def new generation total 18432K, used 3703K [0x045e0000, 0x059e0000, 0x059e0000) eden space 16384K, 22% used [0x045e0000, 0x0497dde0, 0x055e0000) from space 2048K, 0% used [0x055e0000, 0x055e0000, 0x057e0000) to space 2048K, 0% used [0x057e0000, 0x057e0000, 0x059e0000) tenured generation total 40960K, used 40959K [0x059e0000, 0x081e0000, 0x081e0000) the space 40960K, 99% used [0x059e0000, 0x081dfff8, 0x081e0000, 0x081e0000) compacting perm gen total 12288K, used 2083K [0x081e0000, 0x08de0000, 0x10de0000) the space 12288K, 16% used [0x081e0000, 0x083e8c50, 0x083e8e00, 0x08de0000)No shared spaces configured. OK let us quickly identify it using JProfile Download JProfile in here  Run JProfile and attach MemoryTest get largest size of  Objects in Memory View in here is Integer then select Integer and go to Heap Walker. get GC Graph for this object  Then you get detail code raise this issue quickly now.  That is enjoy it.

    Read the article

  • Parameters for selection of Operating system, memory and processor for embedded system ?

    - by James
    I am developing an embedded real time system software (in C language). I have designed the s/w architecture - we know various objects required, interactions required between various objects and IPC communication between tasks. Based on this information, i need to decide on the operating system(RTOS), microprocessor and memory size requirements. (Most likely i would be using Quadros, as it has been suggested by the client based on their prior experience in similar projects) But i am confused about which one to begin with, since choice of one could impact the selection of other. Could you also guide me on parameters to consider to estimate the memory requirements from the s/w design (lower limit and upper limit of memory requirement) ? (Cost of the component(s) could be ignored for this evaluation)

    Read the article

  • Cocos2d update leaking memory

    - by Andrey Chernukha
    I have a weird issue - my app is leaking memory on device only, not on a simulator. It is leaking if i schedule update method anywhere, on any scene. It is leaking despite update method is empty, there's nothing inside it except NSLog. How can it be? I have even scheduled update on the very first scene where it seems there's nothing to leak, and scheduled another empty and it's leaking or not leaking but allocating something, the result is the same - the volume of the memory consumed is increasing and my app is crashing soon. I can detect the leakage via using Instruments-Memory-Activity Monitor or with help of following function: void report_memory(void) { struct task_basic_info info; mach_msg_type_number_t size = sizeof(info); kern_return_t kerr = task_info(mach_task_self(), TASK_BASIC_INFO, (task_info_t)&info, &size); if( kerr == KERN_SUCCESS ) { NSLog(@"Memory in use (in bytes): %u", info.resident_size); } else { NSLog(@"Error with task_info(): %s", mach_error_string(kerr)); } } Can anyone explain me what's going on?

    Read the article

  • Good free software to track and graph memory and CPU usage on Windows?

    - by TheLQ
    There are many questions on Linux memory tracking, but I haven't seen any for Windows. In my case however its a Windows XP Pro box I need to track the memory and CPU usage of. The reason I need it is due to a server program I'm trying that is eating all my processor and some of my memory which is freezing my RDP session and System Explorer and even makes it difficult to login physically. As this is a very constrained server I'm working off of (768 MB RAM with Pentium 4 which disappears with this program), I need a program that doesn't run/require a webserver. I can give it a MySQL database if necessary however. Is there any suggestions for such a program?

    Read the article

  • Does lshw list the "factory" speed of a memory module or the effective speed and how to find the former?

    - by Panayiotis Karabassis
    I hope I phrased this correctly. lshw gives: description: DIMM Synchronous 400 MHz (2.5 ns) product: M378B5773CH0-CH9 vendor: Samsung physical id: 0 slot: DIMM0 size: 2GiB width: 64 bits clock: 400MHz (2.5ns) And indeed the memory speed is set is set to 800MHz in the BIOS, which I think makes sense since it is a double rate. On the other hand, Googling strongly suggests that to this product number corresponds the PC3-10600 type, which is 1333MHz, not 800MHz. And this seems to be confirmed in the BIOS, where if I select Auto for memory bus speed, 1333MHz is selected "based on SPD settings". However in the latter case, the computer does not boot, i.e. the kernel panics, complaining that something attempted to kill the Idle process. So, I am I am beginning to suspect that I have been given defective memory, the technician that installed saw this, and lowered the bus speed. Is this a possibility?

    Read the article

  • Faster Memory Allocation Using vmtasks

    - by Steve Sistare
    You may have noticed a new system process called "vmtasks" on Solaris 11 systems: % pgrep vmtasks 8 % prstat -p 8 PID USERNAME SIZE RSS STATE PRI NICE TIME CPU PROCESS/NLWP 8 root 0K 0K sleep 99 -20 9:10:59 0.0% vmtasks/32 What is vmtasks, and why should you care? In a nutshell, vmtasks accelerates creation, locking, and destruction of pages in shared memory segments. This is particularly helpful for locked memory, as creating a page of physical memory is much more expensive than creating a page of virtual memory. For example, an ISM segment (shmflag & SHM_SHARE_MMU) is locked in memory on the first shmat() call, and a DISM segment (shmflg & SHM_PAGEABLE) is locked using mlock() or memcntl(). Segment operations such as creation and locking are typically single threaded, performed by the thread making the system call. In many applications, the size of a shared memory segment is a large fraction of total physical memory, and the single-threaded initialization is a scalability bottleneck which increases application startup time. To break the bottleneck, we apply parallel processing, harnessing the power of the additional CPUs that are always present on modern platforms. For sufficiently large segments, as many of 16 threads of vmtasks are employed to assist an application thread during creation, locking, and destruction operations. The segment is implicitly divided at page boundaries, and each thread is given a chunk of pages to process. The per-page processing time can vary, so for dynamic load balancing, the number of chunks is greater than the number of threads, and threads grab chunks dynamically as they finish their work. Because the threads modify a single application address space in compressed time interval, contention on locks protecting VM data structures locks was a problem, and we had to re-scale a number of VM locks to get good parallel efficiency. The vmtasks process has 1 thread per CPU and may accelerate multiple segment operations simultaneously, but each operation gets at most 16 helper threads to avoid monopolizing CPU resources. We may reconsider this limit in the future. Acceleration using vmtasks is enabled out of the box, with no tuning required, and works for all Solaris platform architectures (SPARC sun4u, SPARC sun4v, x86). The following tables show the time to create + lock + destroy a large segment, normalized as milliseconds per gigabyte, before and after the introduction of vmtasks: ISM system ncpu before after speedup ------ ---- ------ ----- ------- x4600 32 1386 245 6X X7560 64 1016 153 7X M9000 512 1196 206 6X T5240 128 2506 234 11X T4-2 128 1197 107 11x DISM system ncpu before after speedup ------ ---- ------ ----- ------- x4600 32 1582 265 6X X7560 64 1116 158 7X M9000 512 1165 152 8X T5240 128 2796 198 14X (I am missing the data for T4 DISM, for no good reason; it works fine). The following table separates the creation and destruction times: ISM, T4-2 before after ------ ----- create 702 64 destroy 495 43 To put this in perspective, consider creating a 512 GB ISM segment on T4-2. Creating the segment would take 6 minutes with the old code, and only 33 seconds with the new. If this is your Oracle SGA, you save over 5 minutes when starting the database, and you also save when shutting it down prior to a restart. Those minutes go directly to your bottom line for service availability.

    Read the article

  • SQL Monitor Custom Metric: Out of memory errors

    The number of out of memory errors that have occurred within a rolling five minute window. If you just want to keep an eye out for any memory errors, you can watch the ring buffers for the Out of memory errors alert when it gets registered there. Get alerts within 15 seconds of SQL Server issuesSQL Monitor checks performance data every 15 seconds, so you can fix issues before your users even notice them. Start monitoring with a free trial.

    Read the article

  • How to Increase Memory Allocated to IIS .NET Application?

    - by Mark Hansen
    We are using Windows 2008 R2 and IIS 7 running on Amazon EC2. IIS is running a single .NET application written in C#. We are having performance issues and I want to give the application more memory, but I cannot figure out how to do it. How do I control the amount of memory that the CLR gets? I'm a total newbie with IIS, .NET and the CLR. If I were working with Java, I would just use the -Xmx flag to increase the memory available to the JVM (e.g., -Xmx3000m for 3GB). But, I cannot seem to figure out how to do this in the Windows world.

    Read the article

  • Reading the memory of a N64

    - by toazron1
    I'm looking for a way to read the memory of a N64, while the game is running, in real time. I have a c# program which hooks into the memory of a running emulator and tracks SSB64 stats. I want to do the same thing with the physical N64. Currently it is possible to read the memory with a gameshark pro, however it's extremely slow, buggy, and not practical for what I am trying to accomplish. Would it be possible to tap into the gameshark, or the N64 directly, to access the memory in real time? Thanks!

    Read the article

< Previous Page | 46 47 48 49 50 51 52 53 54 55 56 57  | Next Page >