As you know snapmirrors are setup from the destination end. Whether through the CLI or the SystemManager GUI, you are working with the destination cluster.
"Cluster unreachable" comes up from the destination side when creating a SnapMirror relationship because the source cluster is unreachable. That's literally it. Can happen in the CLI and the GUI. I do most of my bulk SM creation using CLI as well, but occasionally use SystemManager for one offs. I've had the error from both sides previously and it always comes back to an IC-LIF, Peering, or rarely a netwok issue (once had some underlying routing issues in the WAN).
For the one you are trying to create in SystemManager - were you able to create the same relationship in CLI?
I am sorry, I did not make myself clear. Let me try again:
1. I am creating volume SnapMirror, not using SnapManager.
2. I am doing so from Source cluster GUI. When I clieck on Mirror, it bring me to next page, requires me to enter the new volume name on DEST cluster, and all other SM configurations. Before that, I need to enter authentication information to connect the DR cluster. I am getting "Cluster unreachable", after I enter them.
I get you - for clarity...the "Cluster GUI" (assuming you are at 8.3?) is what used to be called "System Manager" which was a separate installable GUI on a workstation. It was integrated into the DoT as a GUI hosted on the cluster itself with 8.3. I'm not referring to "SnapManager" when I say "System Manager". This should get our terminology consistent with each other.
Question remains - if you create this SnapMirror relationship using the CLI rather than the GUI - does it work or do you get the same issue? I'm just curious about that.