[Moabusers] Moab 5.0.0p1 SEGV's on Opterons

David B Jackson jacksond at clusterresources.com
Wed Jan 17 19:39:55 MST 2007


Chris,

  I think we have it fixed in the latest snap release but to confirm, can
you let us see your resource management config, in particular, what
native interfaces do you have defined and is it possible one or more of
them are mal-formed, ie they do not point to real files?

Thanks,
Dave

> Moab 5.0.0p1 and later snapshots seems to break rather badly on Opterons
> (not
> tried it on anything else yet).  This is on FC5.
>
> ==17460== Process terminating with default action of signal 11 (SIGSEGV):
> dumping core
> ==17460==  Access not within mapped region at address 0x0
> ==17460==    at 0x12092730: strcpy (in /lib64/libc-2.4.so)
> ==17460==    by 0x51B3A4: MNatClusterQuery (MNatI.c:1207)
> ==17460==    by 0x5FB045: __MUTFunc (MUtil.c:7424)
> ==17460==    by 0x5FAF55: MUThread (MUtil.c:7397)
> ==17460==    by 0x4BC832: MRMClusterQuery (MRM.c:1144)
> ==17460==    by 0x4BBE54: MRMUpdate (MRM.c:795)
> ==17460==    by 0x47681F: MSchedProcessJobs (MSched.c:7420)
> ==17460==    by 0x505313: MSysMainLoop (MSys.c:10553)
> ==17460==    by 0x404699: main (MServer.c:320)
>
> :-(
>
> --
>  Christopher Samuel - (03)9925 4751 - VPAC Deputy Systems Manager
>  Victorian Partnership for Advanced Computing http://www.vpac.org/
>  Bldg 91, 110 Victoria Street, Carlton South, VIC 3053, Australia
>
> _______________________________________________
> moabusers mailing list
> moabusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/moabusers
>



More information about the moabusers mailing list