Thank you all for your comments and suggestions. I am very sorry not to have responded before now. Last week I was preoccupied with the problem I described in the following thread and then went away for a long weekend. http://www.gluster.org/pipermail/gluster-users/2012-October/034610.html The excessive load problem turned out to be caused by the problematic upgrade and subsequent downgrade that I reported in this thread. I suspect this was because I downgraded using the 3.3.0 packages I found at bits.gluster.com rather than the ones in Kaleb's repo. Unfortunately I have not managed to reproduce the CARP problem on upgrade using a pair of test servers. Therefore the NFS crashing problem with CARP I experienced must have been caused by something unique to my production cluster. Two things I didn't try while attempting to reproduce the NFS crashes were (1) upgrading from 3.3.0-1 and (2) using a mixture of CentOS-5 and CentOS-6 servers. I am now using version 3.3.0-11, and I will put some more effort into reproducing the CARP problem for a bug report if the same thing happens the next time I try to upgrade my production cluster to 3.3.1. Regards -Dan. On 10/22/2012 03:08 PM, Kaleb S. KEITHLEY wrote: > On 10/22/2012 09:42 AM, Dan Bretherton wrote: >> >> Incidentally, when I decided to downgrade 3.3.0 I discovered that those >> RPMs aren't available for download from http://download.glusterfs.org or >> http://repos.fedorapeople.org/repos/kkeithle/glusterfs (epel-glusterfs) >> any more. > > The old 3.3.0 RPMs from my fedorapeople.org repo are at > http://repos.fedorapeople.org/repos/kkeithle/glusterfs/old/... > > >> I managed to find RPMs for version 3.3.0 by Googling for the >> file names and found them here: >> http://bits.gluster.com/gluster/glusterfs/3.3.0/x86_64/. > > I would not recommend using those. They were built on an ancient > CentOS machine that hasn't been updated since forever and I believe > I've heard that they have obsolete dependencies. YMMV. >
