RE: clustered 2003 file server "stalls" for no apparent reason

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance

Still no soap, folks.

The user-mode performance trap ( 3-second polling and mini-dump on 2gb
server) described in Mike's previous response did not stop the active node,
and didn't produce dump.

Am still trying to find a perfmon that might show me where this "hang"
occurs. Of course, I can't predict the time of occurance, so it'll be dumb
luck if I can actually catch a hang in progress.

Entertaining any ideas about how I might track this one down.

Thanks again.

"Mike Rosado [MSFT]" wrote:

> Just keep the newsgroup posted of your progress with regards the your DNS
> theory, so that others can learn from your experience.
>
> --------------------
> Thanks in advance,
> Mike Rosado
> Windows 2000 MCSE + MCDBA
> Microsoft Enterprise Platform Support
> Windows NT/2000/2003 Cluster Technologies
>
> ====================================================
> When responding to posts, please "Reply to Group" via your newsreader so
> that others may learn and benefit from your issue.
> ====================================================
>
> This posting is provided "AS IS" with no warranties, and confers no rights.
> <http://www.microsoft.com/info/cpyright.htm>
>
> -----Original Message-----
> > From: <marty@xxxxxxxxxxxxxxxxxxxxxxxxx>
> > Subject: RE: clustered 2003 file server "stalls" for no apparent reason
> > Date: Tue, 6 Dec 2005 05:35:05 -0800
> >
> > mike;
> >
> > thanks for your response.
> > i've been reading the "hang" detection reference, and it tastes awfully
> > familiar, so i'll engage that as soon as i'm able.
> >
> > other considerations overnight;
> > since i wasn't getting any indication of failure on the server AND i'm a
> > conspiracy theorist, thinking that perhaps:
> > the DNS might be choking (outside my administrative purview), so clients
> > might be unable to resolve to the host IP
> >
> > again, sincerely appreciate your time.
> >
> > "Mike Rosado [MSFT]" wrote:
> >
> > > Hi Marty,
> > >
> > > Have you tried setting it up User Mode Hang Detection to force a memory
> > > dump next time it hangs? Then you can open a case with Microsoft
> Product
> > > Support Services to get some to loan and load the memory dump for root
> > > cause analysis.
> > >
> > > 815267 How to enable User Mode Hang Detection on a server cluster in
> Windows
> > > http://support.microsoft.com/?id=815267
> > >
> > > --------------------
> > > Hope this helps,
> > > Mike Rosado
> > > Windows 2000 MCSE + MCDBA
> > > Microsoft Enterprise Platform Support
> > > Windows NT/2000/2003 Cluster Technologies
> > >
> > > ====================================================
> > > When responding to posts, please "Reply to Group" via your newsreader so
> > > that others may learn and benefit from your issue.
> > > ====================================================
> > >
> > > This posting is provided "AS IS" with no warranties, and confers no
> rights.
> > > <http://www.microsoft.com/info/cpyright.htm>
> > >
> > > -----Original Message-----
> > > > From: <marty@xxxxxxxxxxxxxxxxxxxxxxxxx>
> > > > Subject: clustered 2003 file server "stalls" for no apparent reason
> > > > Date: Mon, 5 Dec 2005 17:21:01 -0800
> > > >
> > > > greetings, and thank you for your attention.
> > > >
> > > > active/passive file service cluster publishes about 6tb to some 3000
> > > users.
> > > > had 2 events recently, where, each time, about 200k files were added
> from
> > > a
> > > > novell file server. when the users were switched to the windows file
> > > services
> > > > (monday morn), for about 4 hours the server stalls about 5-25 times
> per
> > > hour,
> > > > but seldom loses client connections. as we transition into late
> tuesday,
> > > the
> > > > stalls reduce in frequency and severity, finally abating about
> wednesday
> > > > morn. active node is dual HT processor with 2gb ram and single gb nic
> for
> > > > public interface. i've tried to instrument processor (few 80%
> spikes),
> > > > server\work item shortages (=0), sorver work queues\available
> threads(3-9
> > > per
> > > > proc), and queue length (few spikes <10). have noticed about 980 open
> > > files
> > > > (many with 0 connections)
> > > >
> > > > much appreciation and respect for anyone that can help me figure this
> one
> > > > out!!
> > > >
> > >
> > >
> >
>
>
.

Relevant Pages

  • Re: My WOTT - Solo Training
    ... A snapping punch with no push isn't going to cause any real damage. ... You don't need an opponent to see it. ... how to control the adrenalin dump, ... It occurs as a natural response to danger or stress - it's part of the ...
    (rec.martial-arts)
  • ISDN / CAPI tool vendors
    ... The CapiPro VCL version does not hang, and I have been able to get _one_ ... of beginner question to their Technical staff:(My colleague subscribed ... to their Mailing List some days ago, but this far no response has come, ... nor link to get to old Mail archives or anything. ...
    (borland.public.delphi.thirdpartytools.general)
  • Re: MVS 4 minute outage
    ... I would propose that this is not a hang but a perceived hang. ... IBM asked for a dump, ... I've asked the programmer NOT submit the possible causing batch job ... Search the archives at http://bama.ua.edu/archives/ibm-main.html ...
    (bit.listserv.ibm-main)
  • Re: True Strange Minor Confrontation
    ... > Him "Fuck you.I should kick your ass.Whyd you dump yopur trash in my ... I dont dump trash out my car window.Maybe it blew ... > other times just to hang up. ...
    (rec.martial-arts)
  • Re: MVS 4 minute outage
    ... Had another hang today - 6 minutes 50 seconds. ... IBM asked for a dump, ... For IBM-MAIN subscribe / signoff / archive access instructions, ... send email to listserv@xxxxxxxxxxx with the message: GET IBM-MAIN INFO ...
    (bit.listserv.ibm-main)