Failover group configuration on a CDOT 8.3 cluster.

[NairAJ]

All, The following shows our current setup of failover groups in our cluster1 and have an issue with failover.

nas001::> failover-groups show

nas001

                                       1G

                                                                       nas001a:a0a, nas001a:a0b                                                

                                                                       nas001b:a0a, nas001b:a0b

The home port for data lif is given as nas001a:a0a

we made the data lif move to nas001b:a0b and took out the cables which essentially makes "nas001b:a0b" down and we lost the connectivity to the data lif.

This should not be occuring because nas001a:a0a, nas001a:a0b and nas001b:a0a were up and running but it did not failed over. 

Any suggestion or are we missing soemthing.

Thanks,

Aj

[aborzenkov]

What is failover policy? Failover group only defines possible candidates; how these candidates are selected is determined by failover policy.

[NairAJ]

This is our first cluster and here is the policy.

nas001::> network interface show -failover

 svm001a

                svm001a_cifs_lif1    nas001a:a0a               system-defined 1G

                        Failover Targets:  nas001a:a0a, nas001a:a0b,

                                                        nas001b:a0a, nas001b:a0b

In this situation at least the data lif should have failed over to the failover group nas001b:a0a when a0b on the same node was not availble.

Thanks.