Set timeout for APD/PDL in vCenter 5.0U1
We are moving to a new SAN that has SAN to SAN replication. Currently we have this working with no issues.
We are now trying to test a failover to the standby SAN. The way we are testing is by shutting down the ports on the switch that the iSCSI are connected too (trying to simulate a complete loss of the production SAN).
Once we shut down those ports, the ESXi hosts of course panic and try to reconnect. I am trying to find a way to tell the hosts to timeout quicker and not to try and reconnect. Since we are on version 5.0 I do not see the APDTimeout option under Advanced Settings --> Misc.
It seems like the hosts will try for 3- minutes or more to reconnect to the LUNS. Because of this, the hosts do not respond to any other commands.
The SAN manufacturer says we need to add the IP of the standby iSCSI SAN and then rescan for the standby LUNS. Once we do this they should pop up and we can add the VM's to the inventory. Not the best way but for a standby i guess it will do. However, we can't get to that point since the hosts keep trying to find the production SAN.
Is their a way to tell the hosts not to keep scanning for the failed SAN until we get it back up?
We are now trying to test a failover to the standby SAN. The way we are testing is by shutting down the ports on the switch that the iSCSI are connected too (trying to simulate a complete loss of the production SAN).
Once we shut down those ports, the ESXi hosts of course panic and try to reconnect. I am trying to find a way to tell the hosts to timeout quicker and not to try and reconnect. Since we are on version 5.0 I do not see the APDTimeout option under Advanced Settings --> Misc.
It seems like the hosts will try for 3- minutes or more to reconnect to the LUNS. Because of this, the hosts do not respond to any other commands.
The SAN manufacturer says we need to add the IP of the standby iSCSI SAN and then rescan for the standby LUNS. Once we do this they should pop up and we can add the VM's to the inventory. Not the best way but for a standby i guess it will do. However, we can't get to that point since the hosts keep trying to find the production SAN.
Is their a way to tell the hosts not to keep scanning for the failed SAN until we get it back up?
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
Thank you. I also spoke with VMWare support and they confirmed that it's only available with 5.1.
I would suggest contacting your storage vendor for their recommendations