SUMMARY: CPU utilization gives spikes in every 5 min.

From: Ashish Nabira (ashishn_at_cadence.com)
Date: 02/26/04

  • Next message: minoti.koul_at_wipro.com: "unable to see the from field."
    Date: Thu, 26 Feb 2004 16:01:34 +0530
    To: <sunmanagers@sunmanagers.org>
    
    

    Hi All;

    Thanks a lot for all who replied. Finally the problem was fixed by
    applying latest patch.

    I am sorry for delay in posting as this machine was huge and I took
    these many days to actually apply the solution.

    Good day.

    Thanks & Regards;
    Ashish Nabira
    " Work is Worship."
    Ph: +91 120 2562842 Etxn:4877
    Mobile: +91 9811907768

    -----Original Message-----
    From: Gregory Shaw [mailto:shawga@unix.stortek.com]
    Sent: Tuesday, February 17, 2004 1:45 AM
    To: Ashish Nabira
    Subject: Re: CPU utilization gives spikes in every 5 min.

    Look into your cron entries to be sure that some clever user isn't
    doing a cron job every 5 minutes.

    On Feb 16, 2004, at 7:32 AM, Ashish Nabira wrote:

    > Hello Gurus;
    >
    > I have a sun machine having 12 processors, 53GB of RAM. It's giving
    > spikes in CPU utilization in every 4-5 min. I also observed that ,
    this
    > behaviour increases in 15-20 days after rebooting and goes upto 90-95
    %
    > utilization spikes. It's mainly "sys" , i.e kernel which is consuming
    > the CPU cycles. Is it related to OS misbehaving?
    > Help is appreciated and I will summarize.
    >
    > Here is the sar o/p
    >
    > # sar 10 50
    >
    > SunOS cds11744 5.7 Generic_106541-17 sun4u 02/16/04
    >
    > 04:35:06 %usr %sys %wio %idle
    > 04:35:16 0 2 0 98
    > 04:35:26 0 1 0 99
    > 04:35:36 0 1 0 99
    > 04:35:47 0 1 0 98
    > 04:35:57 0 1 0 99
    > 04:36:07 0 1 0 99
    > 04:36:17 0 2 0 98
    > 04:36:27 0 1 0 99
    > 04:36:37 0 1 0 99
    > 04:36:47 0 1 0 98
    > 04:36:57 0 1 0 99
    > 04:37:07 0 1 0 99
    > 04:37:17 0 2 0 98
    > 04:37:27 1 36 0 63
    > 04:37:37 0 44 0 56
    > 04:37:47 0 45 0 55
    > 04:37:57 0 44 0 56
    > 04:38:07 0 2 0 97
    > 04:38:17 0 2 0 98
    > 04:38:27 0 1 0 99
    > 04:38:37 0 1 0 99
    > 04:38:47 0 2 0 99
    > 04:38:57 0 1 0 99
    > 04:39:07 0 1 0 99
    > 04:39:17 0 2 0 98
    > 04:39:27 0 1 0 99
    > 04:39:37 0 1 0 99
    > 04:39:47 0 1 0 99
    > 04:39:57 0 1 0 99
    > 04:40:07 0 2 0 98
    > 04:40:17 0 2 0 98
    > 04:40:27 0 1 0 98
    > 04:40:37 0 2 0 98
    > 04:40:47 0 2 0 98
    > 04:40:57 0 1 0 98
    > 04:41:07 0 2 0 98
    > 04:41:17 0 2 0 98
    > 04:41:27 0 1 0 99
    > 04:41:37 0 1 0 99
    > 04:41:47 0 2 0 98
    > 04:41:57 0 1 0 99
    > 04:42:07 0 1 0 99
    > 04:42:17 0 2 0 98
    > 04:42:27 1 38 0 61
    > 04:42:37 0 43 0 57
    > 04:42:47 0 44 0 56
    > 04:42:57 0 43 0 57
    > 04:43:07 0 2 0 98
    > 04:43:17 0 2 0 98
    > 04:43:27 0 1 0 99
    >
    > Average 0 8 0 92
    >
    >
    > # top
    >
    > load averages: 6.32, 3.00, 2.23
    > 06:22:38
    > 94 processes: 90 sleeping, 1 zombie, 3 on cpu
    > CPU states: 51.2% idle, 0.1% user, 48.7% kernel, 0.0% iowait, 0.0%
    > swap
    > Memory: 56G real, 44G free, 136M swap in use, 123G swap free
    >
    > PID USERNAME THR PRI NICE SIZE RES STATE TIME CPU COMMAND
    > 26135 root 1 0 0 2448K 2064K cpu14 1:55 3.92% top
    > 26545 root 1 0 0 936K 720K cpu22 0:00 0.09% df
    > 923 root 1 58 0 9408K 5000K sleep 45:18 0.03% scopeux
    > 26405 root 4 48 0 3864K 3240K sleep 0:00 0.03% perl
    > 1036 root 12 58 0 2848K 2536K cpu4 5:09 0.01% mibiisa
    > 14673 root 1 59 0 1920K 1400K sleep 0:35 0.01% rpc.rstatd
    > 1090 root 11 20 0 12M 6288K sleep 0:34 0.01% rep_server
    > 795 root 1 58 0 5624K 4728K sleep 0:31 0.00% jre
    > 806 root 1 59 0 6848K 5352K sleep 0:15 0.00% jre
    > 14758 gottsch 1 58 0 5688K 4936K sleep 0:09 0.00% Xvnc
    > 26383 root 1 40 0 1792K 1352K sleep 0:00 0.00%
    > pwc_fsmon.sh
    > 26544 root 1 48 0 992K 928K sleep 0:00 0.00% df
    > 673 root 6 58 0 7720K 7176K sleep 8:10 0.00% automountd
    > 868 root 5 58 0 3064K 2440K sleep 2:16 0.00% mountd
    > 814 root 1 59 0 2344K 1704K sleep 1:31 0.00% nmbd
    >
    > ===============================================
    >
    >
    > Thanks & Regards;
    > Ashish Nabira
    > " Work is Worship."
    > Ph: +91 120 2562842 Etxn:4877
    > Mobile: +91 9811907768
    > _______________________________________________
    > sunmanagers mailing list
    > sunmanagers@sunmanagers.org
    > http://www.sunmanagers.org/mailman/listinfo/sunmanagers
    >
    -----
    Greg Shaw Phone: (303) 673-8273
    IT Architect, Information Systems Fax: (303) 673-2733
    Storage Technology Corporation
    1 StorageTek Drive MS 4382 shawga@vinson.stortek.com (work)
    Louisville CO 80028-4382 shaw@fmsoft.com (home)
    "When Microsoft writes an application for Linux, I've Won." - Linus
    Torvalds
    PGP Fingerprint: F5 9B 8E 1A 9E 21 BA CF 9C D8 F3 7A 17 47 C9 A1
    _______________________________________________
    sunmanagers mailing list
    sunmanagers@sunmanagers.org
    http://www.sunmanagers.org/mailman/listinfo/sunmanagers


  • Next message: minoti.koul_at_wipro.com: "unable to see the from field."

    Relevant Pages

    • Re: Greg Kemnitz PLEASE contact me.............!!!!!!!!!!!!!!!!!!!!!!!
      ... I don't read the newsgroup very often, so I did not see this posting ... packed in the exact packaging that you ... driving about 40 miles to where I have some stuff stored and emptying ... Hopefully your cpu case and contents have arrived by now. ...
      (rec.games.pinball)
    • X-Box 360 Defective Products
      ... First off this information I'm posting for first time buyers. ... Most of you are not aware that X-Box 360 is based on a custom IBM ... PowerPC-based "Xenon" central processing unit (CPU) and a custom ATI ...
      (alt.games.video.xbox)
    • Re: Tech: Stargate Boot-up Problem(s)
      ... I have replaced the U6 and U11 on ... WTB posting turned up nothing. ... the game on without the ribbon cable between the CPU board and the ... I get healthy LED flashes on the boards... ...
      (rec.games.pinball)
    • Re: Directx 9.0bcrash problem with sapphire 9600xt
      ... This posting is provided "AS IS" with no warranties and confers no rights. ... > was set up to write to the location you described but no minidump ... > hopefully this should solve the problem, unless the cpu needs ... power. ...
      (microsoft.public.windowsxp.games)
    • SDSF CPU% vs. ECPU% [re-sent as plain text]
      ... CPU% is time spent "on behalf of" the address space ... JOBNAME CPU% CPU-Time SCPU% Workload SrvClass ECPU-Time ECPU% ... and David in the previous posting) because CPU includes that same enclave CPU. ...
      (bit.listserv.ibm-main)