Concurrent GC threads high CPU utilisation

Hello,
We are running Java 1.5.0_07 with "-XX:MaxPermSize=256M XX:+UseConcMarkSweepGC -XX:+OverrideDefaultLibthread" on Solaris 2.8, 4xSPARC CPU, 8GB memory box.
We get random problems with the GC. Sometimes the 4 threads named: "Gang worker#N (Parallel GC Threads)" go into a CPU frenzy (the thread numbers are always 0, 1, 2 and 3, mapped to LWPs 2, 3, 4, 5). When this happens, each of these threads chews up 16-22% CPU. Overall, the java process uses 65-95% CPU. It lasts for about an hour, and drops back to normal, which is about 1-3% for each of these threads, and 8-15% for the whole process.
This hits about once a week, but it is unpredictable. When it hits, it causes trouble for other system that run on the same box.
Does anyone have a clue on this?
Any help would be appreciated.
Laszlo

Thanks Ramki,
We have found that the default space allocated in the heap for the young generation was much too small. We use a 3 gig heap, on which only 16 meg was allocated to the new generation (all of this went into the eden space - no space could be allocated to the survivor spaces).
Increasing the size of the new generation (now 700 meg - we still may tweak this figure) vastly improved performance. Also the number of garbage collections executed decreased dramatically.

Similar Messages

  • Taskhost.exe DfpCommon.dll thread high CPU utilization

    Pic: http://fud.community.services.support.microsoft.com/Fud/FileDownloadHandler.ashx?fid=f1b9ab2d-9b4e-4ed9-ba70-21a1e1d8e304
    Looks like a bug. What's happening?
    Stack:
    ntoskrnl.exe!KeSynchronizeExecution+0x2246
    ntoskrnl.exe!KeRemoveQueueEx+0x108e
    ntoskrnl.exe!KeRemoveQueueEx+0xae9
    ntoskrnl.exe!KeWaitForSingleObject+0x22a
    ntoskrnl.exe!KeSetBasePriorityThread+0x4ec
    ntoskrnl.exe!KeRemoveQueueEx+0x281d
    ntoskrnl.exe!KeSynchronizeExecution+0x4133
    DfpCommon.dll!IsRunningElevated+0x5e98
    DfpCommon.dll!DllUnregisterServer+0xda7e
    DfpCommon.dll!DllUnregisterServer+0xe49d
    DfpCommon.dll!DllUnregisterServer+0xedc4
    DfpCommon.dll!DllUnregisterServer+0xeafe
    DfpCommon.dll!DllUnregisterServer+0x161e5
    DfpCommon.dll!DllUnregisterServer+0x1a8a0
    DfpCommon.dll!DllUnregisterServer+0x1a9a7
    DfpCommon.dll!DllUnregisterServer+0x1a7fb
    DfpCommon.dll!DllUnregisterServer+0x19782
    DfpCommon.dll!DllUnregisterServer+0x1c537
    DfpCommon.dll!DllUnregisterServer+0x1f340
    DfpCommon.dll!DllUnregisterServer+0x1eb24
    DfpCommon.dll!IsRunningElevated+0x17e3c
    DfpCommon.dll!IsRunningElevated+0x17ff6
    KERNEL32.DLL!BaseThreadInitThunk+0xd
    ntdll.dll!RtlUserThreadStart+0x1d

    Checked this ?
    http://answers.microsoft.com/en-us/windows/forum/windows8_1-performance/taskhostexe-dfpcommondll-thread-high-cpu/c16ce12f-5bc4-4f59-9e08-7404c554a8e5
    Arnav Sharma | http://arnavsharma.net/ Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading
    the thread.

  • How to restart IIS on high CPU utilisation

    Hi There,
    Heres some background on a problem I'm trying to mitigate with SCOM 2010.
    Got a problem with a web server, occasionally the web app causes the CPU to max out and stay there. We can't pick when it happens and often it happens out of hours when everyone is asleep so we don't know that the CPU is running maxed out until the morning.
    At this stage the only way to resolve CPU utilisation is to do iisreset on the affected server.
    We're still trying to figure out what is causing the problem but until then it would be really good to have SCOM detect the high CPU on the server and restart IIS for me. That way the site performance recovers quicker because we wouldn't need someone to
    notice the alert and issue the iisreset.
    I've been tooling around in SCOM trying to figure out how I can have SCOM restart IIS on the affected server when the CPU spikes but I'm just not sure how to do it.
    I only want this IIS Restart Recovery action to only affect the server that has been playing up, rather than affecting all servers that run IIS.
    It would be good to have SCOM detect that the IIS service (w3wp) is the service maxing out the CPU and to restart it on the affected server. But I'd also accept SCOM simply detecting high CPU (> 95%) on the CPU and restarting the IIS service because nothing
    else on the server has been seen to cause the CPU to max out like IIS does.
    Suggestions very much appreciated.

    Please try if this helps:
    Create a custom rule Performance Based->Windows Performance with Counter as Processor & threshold as 95%. Keep the rule disabled.
    Goto properties of that Rule & under Configuration, create a command Response to restart IIS service.
    Override the rule & enable it only for IIS computer group. 
    Hope it helps !!

  • Rpcbind high CPU utilisation

    Hi,
    The issue is my CPU utilisation is unusually high. I've got a dual core laptop and one core is at 100% all of the time.  When I run top I can see that rpcbind is responsible for the high CPU usage. If I reboot the problem starts immediately.
    I've only noticed this problem quite recently, but can't think of anything I've done that may have caused it.
    This problem is really quite urgent because if I can't get it solved I will have to spend days installing another OS because at the moment my laptop battery is draining really fast.
    Does anyone know what may be causing the problem, or how I can troubleshoot it?
    Thanks in advance.

    I figured out the problem. When I upgraded my kernel the other day it turns out that the kernel was installed on my root partition instead of on my boot partition because /boot was not mounted at the time. As a result my old kernel was still being booted, but parts of the system were affected by the new kernel supposedly being installed. So what I did was mount the /boot partition, reinstall the latest kernel and reboot the system - problem solved - no rpcbind taking up 100% of the CPU now.
    In the end I decided to scrap fam and use gamin anyway, based on the wiki recommendation.

  • High CPU utilisation on 2960X stack

    Hello !
    i have a lot of switches 2960-X stack in my network, but i downgraded the iOS version (15.0(2)EX4 to 15.0(2)EX3 be cause some bugs are on EX4 with SFP).
    Since I downgraded the version, i have a lot of troubles on some switches with a huge amount of CPU utilisation.
    One is "normal" and this is the famous one : Hulc LED
    but the other one is not "normal" for me :
    this line :
    190   207178513     1061976     195092  0.00% 31.06% 24.83%   0 Auth Manager
    i dont know why on some switches the "Auth Manager" CPU utilisation is too high.
    When the CPU is too high, we cannot join the switch, the ping is too high (between 500ms and 2100) and users complaining about network spikes.
    So, i dont want to upgrade the iOS to the EX5 (too recent, and i dont know if bug with SPF and CPU are fixed).
    About the Auth manager, what could raise this process to high for the CPU ?
    (i have some check with nagios and snmpwalk but every 5 min and not every minutes, and nagios works well with EX4, issues appear since EX3)
    thank for your help and advice

    15.2(2)E is the first main-line engineering software for the 2960X.  The "EX" versions you've been using are interim version and not really meant for long-term production network.  

  • High cpu utilisation while generating HUM reports

    Hi
    i am facing the issue that while generating the HUM reports the CPU utilisation going to 100%. i am using HUM 300 + LMS 100. right now i am using 2GB RAM

    you're right. i forgot that i made some changes to my igoogle page. the chat function is the culprit. adding it causes the spike. i've reported the problem. thanks for the help!

  • Firefox 13 High CPU Utilisation

    Hello I have installed the latest update and have realised that Firefox’s CPU has sky rocketed it sometimes dips to a normal level but the majority of the time it is around 8-13. With just firefox running it is around 20% utilisation. .
    I have tried using safe mode and have used CCleaner to completely remove history, compact database etc. This unfortunately hasn’t improved the problem.
    I would appreciate any advice on how to resolve the issue and wondered if anyone else has had similar issues.
    Thanks
    BTW this is on a Windows 7 x64 machine if that helps?

    Create a new profile as a test to check if your current profile is causing the problems.
    See "Basic Troubleshooting: Make a new profile":
    *https://support.mozilla.org/kb/Basic+Troubleshooting#w_8-make-a-new-profile
    There may be extensions and plugins installed by default in a new profile, so check that in "Tools > Add-ons > Extensions & Plugins" in case there are still problems.
    If the new profile works then you can transfer some files from the old profile to that new profile, but be careful not to copy corrupted files.
    See:
    *http://kb.mozillazine.org/Transferring_data_to_a_new_profile_-_Firefox

  • Solution to High CPU Utilisation With *Nothing* Running After CS4 Install on Mac OS 10.6.1.

    I'm posting this in case someone else has the same problem. (If anyone else even notices the problem.)
    Another possible title [to help with people searching for the solution] "Computer runs slow after installing CS4 on Mac OS 10.6.1 Snow Leopard"
    I just purchased CS4 Design Premium and have been trying to install it on a fresh install of Snow Leopard (10.6.1) and whilst everything appears to be fine, my CPU sits at 60% with *every* application closed. The instant I uninstall CS4 the CPU drops back to 1-2%. I've conducted this test twice already---that is installed CS4 on a fresh Snow Leopard twice over.
    To be clear this isn't resolved by restarting, as is the solution to one known bug mentioned in the CS4 read-me.
    I did eventually solve the problem. The tech support guy at Adobe was useless, even though I told him that I had re-installed twice and once re-installed Snow Leopard, he had me run the Adobe cleaning utility thingy. Three hours of re-installing (for the third time) and another 700MB of updates (from my precious download quota) and I was exactly back to where I started.
    Then when the updates for Version Cue and Adobe Drive failed to install, it got me thinking. Looking at the box it states that Java is needed for Version Cue Server. That being a unique requirement and the update for that particular program failing, sent me on the search. I eventually found the directory containing Version Cue Server and moved it. I next planned to restart but I didn't have to, the instant I moved the directory my CPU usage plummeted from 60% to 1%.
    I've been working on this for days and all I had to do was issue the command.
    mv /Library/Application\ Support/Adobe/Adobe\ Version\ Cue\ CS4/Server ~/Disabled
    *sigh*
    David.

    Thank you for posting this, I just spent the last 72 hours trying to figure out why my macbook was suddenly burning up my lap (and my battery) without any visible processes in Activity Monitor. This is a pretty major bug (and not the first major one by Adobe on mac) but I am just glad that I didn't need to do a full reinstall. Getting rid of Version Cue takes care of the problem straight out. Thanks for pointing me in the right direction, hopefully others will find this before trying to track down other sources.

  • High CPU Usage in Word 2013

    Hi everyone,
    I'm having an issue with Word 2013 and I'm hoping someone can help. Seen as I got fleeced for my TechNet subscription this year (the price given on the invoice doubled when I called up to pay!) I’d very much like to see a speedy fix from the Microsofties
    please! :D
    Recently, I've found that Word will consistently exhibit high CPU utilisation, ~50% on a dual core second generation Core i5.
    This continues, even when I’m not providing any input to or interacting with the Word window and even when the Word window is covered by other windows. Minimising Word causes its CPU usage to drop back down to zero.
    The strange thing is that, so far, I’ve only observed this with a single document. I started writing this document at the start of this week and it’s been authored exclusively on Word 2013. In fact I’m the only one who’s editing it and I’ve only used a single
    computer (and this one single install of Word 2013) to do so…
    There’s nothing particularly unusual about this document, and a different but similar document which uses the same themes and features (in fact the theme for the second document was created from the first one) doesn’t appear to cause the same issues. Both
    these documents are stored on SkyDrive, and are being edited directly from there.
    Initially I tried the much recommended tweaks of disabling hardware graphics acceleration and sub-pixel positioning, however neither of these has had any effect.
    I investigated a bit further using Sysinternals Process Monitor, and found that while Word is using a lot of CPU it’s actually looping very, very quickly polling for the existence of the registry key “HKCU\Software\Microsoft\Avalon.Graphics”.
    This key doesn’t exist on my system, and a quick test I conducted where I created it just lead to Word looping polling for more non-existent registry keys.
    My system is a HP Folio 13 running Windows 8 Pro x64 with Office 2013 (RTM, not Beta) 32 bit. Both Windows and Office are fully patched, as is everything else on the system.
    Thanks in advance for any help anyone can offer.
    Chris

    Hello again,
    After working on my problematic document some more, I believe I’ve found the cause of the issue.
    @Microsofties – I think this one is either a bug or a dodgy OOTB template…
    I noticed that the high CPU usage only seemed to occur when the document’s footer was displayed. Experimenting with some careful scrolling and zooming to hide footer (without actually using the “hide white-space” option) seemed to confirm this so I investigated
    the footer a bit more.
    My document was using the standard Word “Facet (Odd Page)” footer template with a couple of fields added to suit my needs, namely “Publish Date” and a page number.
    The template seems to be structured such that the content of the footer is inside two nested text boxes (one inside the other). I also noticed previously that this footer template also appears to “jump” up each page and back down again as I scrolled.
    So I deleted the footer and its text boxes, and recreated the same content, in the same position (using the standard paragraph alignment tools) by hand without the nested text boxes. This has, so far, fixed the issue!
    I also noticed that the CPU usage dropped right down to idle the moment I deleted the nested text boxes.
    Altogether, this leads me to conclude that the “Facet (Odd Page)” footer template in Word 2013 is the source of my issue. Specifically, I believe that the use of nested text boxes in this template causes Word’s interactions with the .Net 4.5 WPF to go a
    bit nuts, as evidenced by the infinite loop of request for the non-existent “Avalon.Graphics” WPF registry key I mentioned previously.
    It would be nice to see some feedback on this one, Microsoft people, if only to know if this is actually a general issue or is specific to my document!
    Thanks, Chris

  • High cpu usage by query

    We have a table named "tbl_geodata" which has address and latitude and longitude values. We
    also have a "History" table which has only latitude and longitude values including other information. What we need is like following...
    We get a set of records based on a query from "History" (lat long values), say 5000 records
    Now we are using the following formula to calculate address from the "tbl_geodata" for each row
    (5000 rows).
        SELECT top 1 geo_street,geo_town,geo_country,( 3959 acos( cos( radians(History.lat) ) cos( radians(
    gps_latitude ) ) cos( radians( gps_longitude ) - radians(History.long) ) + sin( radians(History.lat) ) sin( radians( gps_latitude ) ) ) ) AS distance FROM tbl_geodata ORDER BY distance
    Currently we are seeing high cpu utilisation and performance issue. What would be the optimized way to do
    this

    We are using SQL Server Web Edition. We have a table which has around 120 million records (every second around 100 insertion). It has a AFTER INSERT trigger to update another table. There is following performance issue we are facing. 
    High CPU Usage
    Insertion failed (Connection Timeout Expired.  The timeout period elapsed during the post-login phase.  The connection could have timed out while waiting for server to complete the login process and respond; Or it could have timed out while attempting
    to create multiple active connections.  The duration spent while attempting to connect to this server was - [Pre-Login] initialization=0; handshake=10046; [Login] initialization=0; authentication=0; [Post-Login] complete=3999;)
    Insertion Failed (A severe error occurred on the current command.  The results, if any, should be discarded.)
    Insertion Failed (Some time we are getting connection pool error.. There is no limit set in the connection string)
    I ran sp_who2 command and found a lot of queries are in suspended mode..
    here is the "sys.dm_os_sys_info" result...
    cpu_count   hyperthread_ratio
    physical_memory_in_bytes virtual_memory_in_bytes
    8                   8
                                12853739520
                 8796092891136
    Can anyone please suggest the improvement steps...

  • High CPU load on 2811

    Hi,
    We have a 2811 router and we are facing high CPU load but show proc cpu does not show processes with high CPU utilisation.
    RT#show proc cpu sorted | exclude 0.00%
    CPU utilization for five seconds: 99%/94%; one minute: 99%; five minutes: 98%
     PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process
      95     2157484   1814314       1189  3.29%  4.76%  4.68%   0 IP Input        
       2       14624     44123        331  0.94%  0.64%  0.63%   0 Load Meter      
       5      154552      2451      63056  0.37%  0.97%  0.91%   0 Pool Manager    
      38      126116    682626        184  0.18%  0.23%  0.24%   0 Net Background  
     185      142660    545202        261  0.09%  0.08%  0.08%   0 Crypto PAS Proc 
      88       58804   6821262          8  0.09%  0.01%  0.01%   0 ACCT Periodic Pr
     102        4548       395      11513  0.09%  0.11%  0.03% 514 SSH Process     
      42       12480    220592         56  0.09%  0.03%  0.02%   0 Per-Second Jobs 
    This does not happed outside working hours. The router is used only for VPN's (No NATing). It terminates around 25 tunnels but there is not a lot of traffic through the tunnels.
    The extremly high load started mondat morning and continues this morning. During the weekend it was fine. Earlier we would have situations when the CPU would spike but it would soon be down to normal values.
    From what I read the above output means that the CPU load is caused by interrupts. What is the best way to troubleshoot this?
    Regards.

    Disclaimer
    The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
    Liability Disclaimer
    In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
    Posting
    Sorry I wasn't clear, but when I was asking about "not a lot of traffic", I was wondering about actual  volume passing through the 2811.
    When I asked about "kind of tunnels", I was wondering about how they are configured.  E.g. GRE or IPSec/GRE or VTI, encryption options, other tunnel interface options.
    Perhaps the easiest way to check for fragmentation, is packet analyze the traffic passing through the 2811.

  • High CPU on CSS11503 , with one Flow

    Hi,
    I've seen some high CPU Utilisation on 2 CSS.
    I've figured out, that this Utilisation was created, by a mirroring Task from a database Server.
    There was one TCP Connection between 2 Servers, whitch transfered about 60 Mbit/s.
    But i saw a CPU-Utilisation from about 30-40% on my SCM which passed the Traffic.
    That is some thing i do not understand, because the Forwading of this Flow shoud be done in the Flow-Processor, and the CPU Utilisation which is monitored, is the Session Processor.
    The Protocol was TCP 1500, which is Used by Tivoli Storage Manager.
    Anybody has some Idea, how this makes so much CPU Utilisation?
    The Transfer was about 4 hours, and the CPU Utilisation also.
    Best regards
    Sven

    Hello Gilles,
    may i try to show you the network design
    <-> <-> <-> css <->
    So the traffic was hitting 2 css.
    One 11506 with 7.50 103
    One 11503 with 7.50.103
    The traffic was one TCP Connection between Server1 and Server2 (TCP 1500).
    There are no ACL and the traffic was not hitting a content rule.
    The 11506 handels normal about 30 mbit/s out, and 10 mbit/s in. Average CPU is about 1% over a day.
    The 11053 handles normaly about 5mit/ out and 1 mbit/s in. Average CPU is about 4% over a day.
    CPU was about 30% higher on both css, during the replicarton (~3 hours)
    I did not took any debug or show commands, because i saw it on my mrtg a day later.
    Also, the high cpu was only seen on the SCM Module, the second Module (IOM) was normal.
    Best regads
    Sven

  • High CPU Utlisation ws2003 and 2008

    I am trying to understand how to analyze High CPU Utilisation in WS 2003 and 2008 .How do I find the process causing the issue and entire days logs
    .Please help me

    Hi Ritesh,
    To troubleshoot the high CPU Usage, you can also check these tools:
    1. Task Manager (built-in to the O.S.)
    2. Process Explorer
    http://technet.microsoft.com/en-us/sysinternals/bb896653.aspx
    Make sure to enable the Microsoft public symbol to get stack information.
    3. Perfmon (built-in to the O.S.)
    4. Windows Performance Toolkit (XPerf is what replaced KernRate for Windows Server 2008 and Windows Server 2008 R2)
    http://msdn.microsoft.com/en-us/performance/default.aspx
    5. Complete memory dump.
    972110 How to generate a kernel dump file or a complete memory dump file in Windows Server 2003
    http://support.microsoft.com/?id=972110
    969028 How to generate a kernel or a complete memory dump file in Windows Server 2008
    http://support.microsoft.com/?id=969028
    And you can also refer to the troubleshooting steps in these article:
    How to troubleshoot high cpu in the System process
    The Case of the System Process CPU Spikes
    If there is anything else regarding this issue, please feel free to post back.
    Best Regards,
    Anna Wang
    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Support, contact [email protected]

  • High number of threads and high CPU usage on a single instance

    Without any apparent reason, in some moments of day the website reports an increment of the HTTP Queue Length from an average of 10/20 queued requests to 100/200 and more.
    In that period of time the Average Response Time increases but the number of the requests on the website don't. Even increasing the instances it doesn't help to resolve the problem (maybe it just reduced it).
    It seems that the problem is not related to the traffic on the site. This issue happens either when the traffic is high or low.
    The autoscaling works well, we don't have peak of traffic but always a slow increasing of requests and the CPU is always below the 50% (memory too). The issue is resolved by the swap of the site and sometimes it's resolve by itself after a while.
    What I discovered when the issue happens, is that there's an instance (just an instance only), that has an high number of threads and CPU usage on the W3WP.exe process and these values are above the mean than the other instances. Eg, instances are around
    50/60 threads and 10/20% of CPU, the instance with the problem has 200 threads and 50/60% of CPU. If I kill the W3WP process on that instance the issue is resolved. 
    I monitored the HTTP requests, I tried the website extension "Diagnostics as service" but I can't discover anything that can help me to understand the problem. Of course, there are many requests that go in timeout but it is a consequences of a
    unresponsive instance.
    On the web hosting plan there are only two sites: the production site and its staging which is used for update of the production only and it's always sleeping. Plus, no webjobs are running on this site.
    What can I do to gather useful information that can help me to understand the reason of this problem?
    Thank you!

    "Does the instance with the high counts receive more traffic than the other instances? Is it possible that the load balancer not
    working the way it should be?"
    How can I get that information? I can't see metrics for a specific instance
    "Does it always happen to the first instance?"
    I will check it better, but in the order given by the Processes panel it's the second instance which has the issue. The scaling is at 2 instances for the most part of the time (it's the minimum).
    Maybe one time it was the 3th instance but I'm not sure, I'll give it more attention.
    " How long do these moment last?"
    The time can be 10 to 30 minutes, but I fix it as soon as I see the problem. It's not the down-scaling to resolve it because in these situations the CPU as well is high so the scaling holds or increases the number of instances.
    "- How often do these moments occurs?"
    It occurs quite often, 2/4 times a week.

  • How to find out the JVM thread that is consuming high CPU

    I am using WL 10.3, JRockit JVM on Linux. I have taken thread dumps. How to find out the JVM thread that is consuming high CPU?
    I know how to map it Sun JVM on Solaris. I haven't done in on JRockit so far. Please advise.

    Check out some of Marcus Hirt's blog entries for details about how to use JRockit Mission Control. It's a very nice tool that should provide a lot of insight:
    http://blogs.oracle.com/hirt/

Maybe you are looking for