Before we switched to xen we used clvm with our fiber scan. You setup clvm the normal way and then just tell cs what disk to use. Fencing was harder than the whole setup.
<div>-------- Original message --------</div><div>From: ilya musayev <ilya.mailing.li...@gmail.com> </div><div>Date:08/20/2014 1:58 AM (GMT-05:00) </div><div>To: users@cloudstack.apache.org </div><div>Subject: Re: Live migration shared storage </div><div> </div>You can try emailing Marcus directly to see if he can help. Regards ilya On 8/19/14, 8:12 PM, John Mancuso wrote: > Would this be a recommended approach? > > http://www.slideshare.net/MarcusLSorensen/cloud-stack-clvm > > > From: John Mancuso > Sent: Monday, August 18, 2014 9:29 PM > To: 'users@cloudstack.apache.org' > Subject: Live migration shared storage > > This may be more of a KVM question but we are evaluating the VM live > migration feature of Cloudstack 4.4 on KVM using shared FC SAN storage on > RHEL 6/7. I'm a bit confused as to the proper setup/best practice for > accomplishing this. For primary storage, we would like to avoid NFS like the > plague as the throughput is much much slower than GFS2, OCFS2, ceph, etc. My > past experience was with Redhat Cluster which can use CLVM with a SAN-backed > ext3 volume to simply deactivate a volume group on a dead node and activate > it on the live one. Would there be anything like that here (Active/passive > volume) or would my best bet be a distributed file system like GFS/OCFS which > I believe is active/active? We only need active/passive to allow failover for > VM obviously. > > According to this we can use Fiber Channel via shared mount, but can someone > explain the best approach? > http://cloudstack.apache.org/docs/en-US/Apache_CloudStack/4.2.0/html/Admin_Guide/primary-storage.html > > > -John > >
