hi, I was wondering if anyone tested the failover of the SAN switch in the EQL SAN environment.
I have 2 Dell 8024F switches and when simulate a fail (unplug one switch), then I lose the connection to the datastores in the SAN, approx half of the VMs become unreachable. The switches are LAGed together and with the Cisco core 6509 switch.
Please let me know if anyone is familiar with the situation and if have any solution.
Thanks,
Haven't seen the first set of errors before, when these errors come up are you seeing connections dropped on the arrays and/or errors logged on the switches?
Are your vmk adapters bound to separate physical adapters which are then connected one each to your 8024Fs? I haven't tested a failure of our iSCSI switches but have experienced the real thing too many times and haven't had any VMs become unreachable.
Check and make sure that you have multiple independent connections between the ESXi hosts, the switches and the arrays; sounds like there is a single path somewhere in the chain.
aydeisen
4 Posts
0
September 16th, 2011 10:00
hi, I was wondering if anyone tested the failover of the SAN switch in the EQL SAN environment.
I have 2 Dell 8024F switches and when simulate a fail (unplug one switch), then I lose the connection to the datastores in the SAN, approx half of the VMs become unreachable. The switches are LAGed together and with the Cisco core 6509 switch.
Please let me know if anyone is familiar with the situation and if have any solution.
Thanks,
MichaelDavis
77 Posts
0
September 16th, 2011 15:00
Are your vmk adapters bound to separate physical adapters which are then connected one each to your 8024Fs? I haven't tested a failure of our iSCSI switches but have experienced the real thing too many times and haven't had any VMs become unreachable.
Check and make sure that you have multiple independent connections between the ESXi hosts, the switches and the arrays; sounds like there is a single path somewhere in the chain.