Firmware 3.x Initiator Group creation using the vCenter plugin

[rvvliet78110]

I noticed that since firmware 3.x when creating a volume using the vCenter plugin something changed with the initiator groups usage.

In the previous firmware when I would create a volume and selected a cluster in the vCenter plugin then the volume would be added to a initiator group that matched the cluster or a new initiator group would be created for this cluster.

This is really nice in case a host is moved to a different cluster, as you only have to edit one initiator group and all volumes in that group would be accessible or inaccessible).

In firmware 3.x I noticed that when creating a new volume a new initiator "group" is created for each individual host in that cluster (even when a initiator group that matching this cluster exists).

This means that one volume is now added to in this case 9 "initiator groups" instead of a single one.

If I now move a host to a different cluster I have to remove this "initiator group" from all volumes and add it to the volumes in the other cluster.

This is a really time consuming business.

I spoke to support and this is by design to match the workflow with iSCSI and FC arrays, but for me this is a real setback leaving me with lots extra work.

Are any of you experiencing the same?

I would like to see the previous workflow back in place or at least a possibility to choose how one would like the workflow to be.

I would also like to know how others think about this?

[mblumberg16 · Accepted solution]

Hi Rick, I understand the time spent over this task and the frustration that you get out of it, apologies for that.

As a side note, we do have an RFE that will allow unmount of a datastore operation via the plugin, and not just deletion.

Such operation can also remove the initiator group from the Nimble volume.

You can use the mount datastore option to mount the datastore on a host in order to avoid the adding of inititator group task.

Additional note: the mapping of 1-1 initiator group and host iqn is also driven by our VVol integration and we had to perform the change.

I hope you find this information useful.

Thanks,

Moshe.

[rvvliet78110]

Hoi Moshe,

Thanks for your reply.

"Additional note: the mapping of 1-1 initiator group and host iqn is also driven by our VVol integration and we had to perform the change."

I didn't know that yet, if this is indeed needed then so be it, as we would like to migrate to vVols then we just have to accept this for now.

Additional question: Do you possibly know if the max volume limit will be upped, as we have about 1500 VM and the current volume limit is 1024 we can't use vVols yet.

Thnx,

Rick

[mblumberg16]

That's correct we're increasing the volume number limit in the next major release, like all vendors that integrate with VMware and VVol, this limitation needed to be removed

Thanks!