[Sunhelp] Sporadic reboots on Sparc 5

[Flynn, Harold M. III]

Send me a snip from /var/adm/messages surrounding the incident, and a
showrev -p.  I've seen a similar incident, and would like to see if there
are any co-relations.

I seem to recall something similar to this from a few months back on the
list, as well.

> -----Original Message-----
> From:	Jim Richey [SMTP:jrichey at highmark.com]
> Sent:	Thursday, January 06, 2000 10:34 AM
> To:	sunhelp at sunhelp.org
> Subject:	Re: [Sunhelp] Sporadic reboots on Sparc 5
> 
> We have the same problem with Solaris 2.6 on UE 3500's. Some systems
> reboot themselves, others just crash. Same situation though; didn't
> start occurring until after applying the Y2K patch cluster. We don't get
> any messages in the logs whatsoever. We just downloaded the latest
> recommended patches and installed them, but there was only one new patch
> and it didn't look like it would fix anything relevant to this problem.
> Please let me know if you find the problem and I'll do likewise.
> 
> S Condit wrote:
> > 
> > Hi All,
> >         I've enjoyed reading the list over the past few months and I've
> learnt
> > quite a few things - hopefully someone can help me out with this one.
> > 
> > We have a sparc 5  that we use for sendmail and apache (mainly). The
> > machine was to be replaced by an Ultra5 before Christmas but the new
> > machine wasn't ready on time. We had to apply the patches to Solaris 2.5
> > (recommended and y2000 - we left out sendmail - openwindows - all the
> stuff
> > we don't need) in a bit of a rush and ever since this the machine has
> been
> > reboting on an irregular basis every 2-6 hours.
> > 
> > /var/adm/messages refers to kernel read and write errors at the time of
> the
> > crash
> > pid varies - sometimes its a httpd process, sometimes sendmail,
> sometimes a
> > system process.
> > It does consistently say 'Invalid Address on supv data fetch'
> > 
> > I've tried reapplying the patches, which didn't help.
> > 
> > Since all the messages refer to bad data there is a possibility that it
> > could be hardware but as the reboots started just after the patches were
> > applied, I would find this unlikely.
> > 
> > Any of you have any thoughts on this - apart from going to the backups
> :-)
> > 
> > TIA
> > 
> > Stephen Condit
> > 
> > _______________________________________________
> > SunHELP maillist  -  SunHELP at sunhelp.org
> > http://www.sunhelp.org/mailman/listinfo/sunhelp
> 
> -- 
> 
> Jim Richey
> jrichey at highmark.com
> Highmark, Inc.
> http://www.highmark.com
> 
>