[DRBD-user] Machine crashed repeatedly: drbd16: Epoch set sizewrong!!found=1061 reported=1060

Todd Denniston Todd.Denniston at ssa.crane.navy.mil
Mon Nov 1 17:14:16 CET 2004


Andreas Hartmann wrote:
> 
> Hello Lars,
> 
> Lars Ellenberg schrieb:
> > sure, it still could be drbd's fault. but there is not real sign of this.
> > and the symptom of "gcc compile fails always the same;
> > after reboot, it works fine" for me clearly suggests bad ram.
> 
> This would mean:
> I've got two machines, both would have bad RAM, which comes up with drbd
> only. All other applications are running fine. Hmmm. Ok, not impossible
> but very hard to believe.
> 
> Could it be to get rid of the potential errors, if the RAM-timings in the
> bios are changed to a less critical value (if this is possible at all)?
> 
> I recall, that there are memory chips (with an additional chip), which can
> detect flipping bits. Don't know, if the machines memory has this feature.
> I will have a look!
> 
I am not sure if the XSeries are x86 machines or not,
if they are you might try memtest86[1], if they are not then the old 'memory
test script'[2] can be useful.  If you have 'upgraded' the ram you might want
to take a quick look at some linuxforum comments[3]. The memtest[2] script
does have one other advantage over the memtest86, you can run it while the
system is live, but it is not necessarily a good idea.  We have found two
systems lately which got a batch of new ram and then would not finish an
install of fedora before crashing, memtest86 showed that a portion of the ram
was failing, we slowed down the ram bus and then it passed memtest86 and
worked, which was acceptable for the system use.


[1] http://linux.maruhn.com/sec/memtest86.html

[2] http://people.redhat.com/dledford/memtest.html

[3] http://www.linuxforum.com/forums/index.php?showtopic=86713

-- 
Todd Denniston
Crane Division, Naval Surface Warfare Center (NSWC Crane) 
Harnessing the Power of Technology for the Warfighter



More information about the drbd-user mailing list