Is the same application run across all of the VMs? A common symptom usually stems from apps that don't behave properly and have a lot of frequent writes. If possible you should spin up a blank compute node and see if the symptoms are similar. > On Oct 15, 2013, at 11:43 AM, Samuel Winchenbach <swinchen at gmail.com> wrote: > > Hi all, > > I have a 16 process/64GB RAM compute node running Grizzly with some odd memory usage behavior. The node is currently running 17 lightly used VMs. > > root at test1:~/scripts# free -m > total used free shared buffers cached > Mem: 64429 48742 15687 0 0 28479 > -/+ buffers/cache: 20263 44166 > Swap: 4093 4026 67 > > As you can see I have over 15GB free (44 if you remove cached memory!) yet the swap space is almost completely full. > > Any idea what might be going on? > > Thanks, > Sam > > > _______________________________________________ > OpenStack-operators mailing list > OpenStack-operators at lists.openstack.org > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20131015/1f823df6/attachment.html>