SUMMARY: SPARC4's freeze in use

From: Chris Stoddart (chris@dcs.shef.ac.uk)
Date: Mon Nov 02 1998 - 10:00:50 CST


Hi managers,

Long time, no summary! Well fingers crossed the solution we have works.
Original message at the bottom of the mail, but the essential fix was to
replace the power supply (the machines are under maintenance, so Sun did
this for free). The engineer suggested if the problem persisted then he'd
swap the motherboard, but so far we have two trouble-free weeks.

Thanks for ideas/suggestions/encouragement to:

Suzanne Taylor <staylor@convergys.com>
Charlie Mengler <charliem@anchorchips.com>
Bill at: Sun System Admin <sysadm@its.brooklyn.cuny.edu>
Frank Smith <franks@tradeonemktg-aus.com>
Michael Neef <Michael.Neef@neuroinformatik.ruhr-uni-bochum.de>
Chris Marble <Chris_Marble@hmc.edu>
Jason Marshall <jasonm@vsl.com>
Rick Schieche <RSCHIEC@co.pierce.wa.us>
Leo Crombach <lcrombach@tropel.com>
Francis Liu <fxl@uts.edu.au>
Michael Hill <Michael_Hill@csgsystems.com>
Ray Brownrigg <Ray.Brownrigg@MCS.VUW.AC.NZ>

Ta muchly,

Chris

--
Dr Chris Stoddart: Unix SysAdmin,
Department of Computer Science, 
Sheffield University, U.K.

On Thu, 8 Oct 1998, Chris Stoddart wrote:

> > Hi guys, > > I have a couple of Sparcstation 4's running Solaris 2.6 that exhibit an > odd lockup when a user is logged in with CDE, olwm or olvwm. Basically the > screen freezes and the machine cannot be pinged or otherwise reached > over the network. More oddly 'stop-a' doesn't work either and the only > recourse is to physically turn the machine off and on again. There are no > messages written to /var/adm, so just little to go on. Also because these > machines are remote from me, I have never seen it happen just the > results. Now *I* have a Sparcstation 4 running Sol 2.6 in exactly the > same configuration, and it has never happened to me and I've been up 6 > months longer than these machines. > > I have tried re-installing Solaris from scratch, the latest patch cluster is > installed, all to no avail. The machines are short on RAM but have more > than enough swap space to compensate (at least 256Mb). > > Has anyone seen this behaviour before? I know it's hard with so little to > go on, but that is what has stumped me! Your ideas gratefully received! > > Chris > >



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:12:51 CDT