Re: AIX Run Queue

From: Hunter, Mark (Mark.Hunter_at_ANHEUSER-BUSCH.COM)
Date: 03/28/05

  • Next message: Andrew Townsend: "Re: SSA vs SCSI - general IO troubleshooting"
    Date:         Sun, 27 Mar 2005 23:29:36 -0600
    To: aix-l@Princeton.EDU
    
    

    I have seen similar condition when I have had database processes spinning on a
    lock. You might check to see if you have an abnormally high
    number of syscalls. Load average of serveral hundred is not normal for a Unix
    box. Given you have idle time, you almost certainly have a resource that
    processes are spinning on.
     
    If possible, you may want to run a system trace. I doubt you can get more than
    a few seconds worth, but it should indicate where the issue may be.
    Mark Hunter
    Anheuser-Busch Cos.
    MIS Consultant, ES&SO Server Planning and Integration
    *Office: (314) 632-6663
    *Fax: (314) 632-6901
    ?Pager: (314) 841-4026
    *Email: Mark.Hunter@Anheuser-Busch.com

    The information transmitted (including attachments) is covered by the Electronic
    Communications Privacy Act, 18 U.S.C. 2510-2521, is intended only for the
    person(s) or entity/entities to which it is addressed and may contain
    confidential and/or privileged material. Any review, retransmission,
    dissemination or other use of, or taking of any action in reliance upon, this
    information by persons or entities other than the intended recipient(s) is
    prohibited. If you received this in error, please contact the sender and delete
    the material from any computer.

     
     
     

      _____

    From: IBM AIX Discussion List [mailto:aix-l@Princeton.EDU] On Behalf Of Green,
    Simon
    Sent: Sunday, March 27, 2005 1:41 PM
    To: aix-l@Princeton.EDU
    Subject: Re: AIX Run Queue

    If performance is acceptable, you don't have a problem! Remember: the objective
    is to satisfy the customers, not to get nice numbers out of the monitoring.
    However, a runqueue as high as you say would worry me somewhat as it usually
    leads to poor performance.
     
    I've seen high rq in the past where an application has had debugging options set
    on. Also when database indices have gone missing, or have to be re-built.
    Otherwise, it's usually just an indication of a very high workload.

    -- 
    Simon Green
    Altria ITSC Europe s.a.r.l.
    AIX-L Archive at https://new-lists.princeton.edu/listserv/aix-l.html
    New to AIX? http://publib-b.boulder.ibm.com/redbooks.nsf/portals/UNIX
     
    N.B. Unsolicited email from vendors will not be appreciated.
    Please post all follow-ups to the list.
            -----Original Message-----
            From: IBM AIX Discussion List [mailto:aix-l@Princeton.EDU] On Behalf Of
    PJ Humble
            Sent: 25 March 2005 16:36
            To: aix-l@Princeton.EDU
            Subject: AIX Run Queue
    	
    	
                    Greetings fellow AIX'ers,
                    Our organization has increased its focus on performance
    monitoring, and capacity planning... The "run queue" factors heavily into the
    generic metrics gathered & monitored (& used for modeling) acrossed the UNIX
    platforms here.
                    Our solaris and hpux servers seem to be able to follow the "cpu
    x3" standard for an acceptable run queue (i.e. on a 2-way, you'd expect your run
    queues to stay below six {6} for optimal performance)...
                    However, on my AIX servers, I routinely see run queues reported
    in the hundreds... I know AIX has a run queue per proc, and reports to a global
    run queue, but can any of the GURUs out there give me an education that'll help
    me demystify why my server's run queues seem frightful in the eyes of other UNIX
    admins...
                    Despite the high reporting, my servers run fine... we're not cpu
    bound (us+sy <80)... I've checked the certification performance guide, and aside
    from a couple cool drawings, I came up empty.
                    Any advice & input appreciated! Thanks in advance!!!
    

  • Next message: Andrew Townsend: "Re: SSA vs SCSI - general IO troubleshooting"

    Relevant Pages

    • PF with 20Mbps data streaming
      ... I'm still working on porting our unicast stream servers behind PF ... # of states in PF and # of connected hosts ... -- QUEUE limits? ... queue internet bandwidth 16Mb ...
      (comp.unix.bsd.openbsd.misc)
    • 20 Mbps data streaming behind PF
      ... I'm still working on porting our unicast stream servers behind PF ... # of states in PF and # of connected hosts ... -- QUEUE limits? ... queue internet bandwidth 16Mb ...
      (comp.os.linux.security)
    • Re: SMTP queue
      ... Those virtual servers are still showing. ... Unable to bind to the destination server in DNS. ... Through the EICW? ... Check the sender/recipient of the messages in the queue to see if they ...
      (microsoft.public.windows.server.sbs)
    • Re: Exchange 2007 CCR - Messages Queued for Submission
      ... Thanks Matt - thats the same solution I had found - failing over the cluster ... seemed to clear this queue. ... Sending/Receiving between all servers works properly.) ...
      (microsoft.public.exchange.admin)
    • Re: To be or not to be
      ... that queue, only to go to sleep, wake up, still be in the queue, only to ... This only happend over a 2 day period when the new Oceanic servers were ... I was quoting Usenet & a number of other web sources ("up to 24 hrs" ... But really Nos I can guarantee that any LoTRO server, ...
      (comp.sys.ibm.pc.games.rpg)