Data Infrastructure Management Software Discussions
Today we saw a lot of servers, which were marked as "blocked host" which means that they are zoned but have not mapped any storage.
When we investigated this, we saw that the servers have mapped storage. The cause is, that OCI mapps the server to the partner site in the MCC and not to the running instance of the SVM.
Is this a known issue?
What protocol are these hosts using to access the MCC? I am not aware of any particular issue here.
We are using FCP.
The LIFs on both sites have the same WWPNs but the LIFs on the standby site are down. OCI seems to use the standby site for checking SAN paths.
Do the LIFS have different WWNNs? OCI uses the node WWN for a fair amount of its logic, so that might be what's behind this one.
No, because of the MCC, the LIFs of the standby SVM has the same WWPNs as the original one. The LIFs of the standby SVM are offline and will be onlined with a metrocluster switchover only.
We now have a patch that should allow OCI to better understand this topology.