On 06/30/2016 02:03 AM, andrei.borzenkov@ts.fujitsu.com wrote:
Yes, it is vague.
As Tim already explained intercluster LIFs are not supposed to failover to another node, and interface group should offer enough redundancy so your configuration looks OK from this side.
I was aware that the IC lifs had to fail over locally. I was just musing if the configuration was causing the warning.
The data lifs which use the ifgroup should and do fail over to the other node but the same ifgroup is used for the IC lifs which fail over to the same lif/ifgroup on the local node.
How does the cluster makes sense of this. Is it possible to "not" set a failover group for the IC lifs given that redundancy exists through the dynamic mode ifgroup if an interface on the node's ifgroup fails during a snapmirror operation.
Regards,
Joel
With best regards
Andrei Borzenkov Senior system engineer FTS WEMEAI RUC RU SC TMS FOS