This post is more than 5 years old
84 Posts
0
6453
SyncRep fail-over behaviour
Setup: 1x Campus, 2x server rooms, direct full mesh fiber between iSCSI switches
In each room the customer has 2 ESXi 5.5 hosts, 2x iSCSI switches and 1xPS6500
The iSCSI network is a flat layer 2 network. The customer wants SyncRep enabled and both server rooms to be active/active to make good ROI. In a DR scenario, for example server room A goes offline. he wants VMs to failover to server room B using VMware HA.
What happens exactly with the SyncRep fail-over. I have read the TR article on this which mentions the naa changes and the restores have to be rescanned. Plus intervention in the Group Manager.
Can anyone elaborate any futher on exactly what to expect and what actions will have to be taken.
Thanks
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
1
February 27th, 2014 09:00
Hello,
The TR is describing what would happen after a failure on the primary side. I.e. a double faulted RAIDset. Since currently there is no automatic failure over to the alternate pool. You need to at GUI/CLI failover to the alternate pool. Then in case of ESXi, rescan and start up your VMs.
If you wanted to do it proactively, you can symply move to the alternate pool w/o any downtime. It's uses a SCSI required features "login redirect" and "Async logout". When you move to the alternate pool, it will log out existing connections and when they immediately relogin they will be redirected to the pool that has the now active sync rep pool.
Regards,