2-Node cluster, Starwind storage not available on either host, Help!

[Fonzies50]

Good morning,
I removed the LUN and rebooted, and everything went fine, it worked right away.
Then I ran a test by disconnecting the network between the two nodes (which, I repeat, is a direct cable), and here the problems began again:
node 1: the partner is not synchronized,
node 2: the current node is not synchronized.
I ran the scripts, but in addition to

Code: Select all

Unable to find the 'MaintenanceMode' property in this object. Check that it exists.
PS C:\software\Starwind> .\MaintenanceMode.ps1
C:\software\Starwind\MaintenanceMode.ps1 : operation cannot be completed. Current or partner node is in a non-synchronized state.
In line:1 char:1
+ .\MaintenanceMode.ps1
+ ~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : NotSpecified: (:) [Write-Error], WriteErrorException
+ FullyQualifiedErrorId : Microsoft.PowerShell.Commands.WriteErrorException,MaintenanceMode.ps1

If I force the node to be synchronized, the two systems don't proceed with the update. The LUN is limited; only connections between the two nodes are active, but only Node 1 is connected to the servers.
I'm starting to get a little frustrated by this situation. I don't understand how this inconsistency can be resolved. Since one node is on one type of hypervisor and the other on another type, I always suspect that having two different licenses means the nodes have differences and are therefore incompatible with each other with advanced option.

[yaroslav (staff)]

Then I ran a test by disconnecting the network between the two nodes (which, I repeat, is a direct cable), and here the problems began again:
node 1: the partner is not synchronized,
node 2: the current node is not synchronized.

This is expected behavior until you reconnect the cable.
See more at viewtopic.php?t=5731
You cannot enter/exit maintenance mode by design. Marking the not syncornized node manually as synchronized is also not a good idea as it overrides the idea behind the priorities.

[Fonzies50]

viewtopic.php?p=32173#p32173

In this file, I only changed the Sync parameter, forcing status 0 on node 2 (the one that actually needs to resync because node 1 is the correct one).
I checked and changed the node priority: should the sync parameter be adjusted accordingly?
Please note that in the log, I never saw an event that said "device synchronized," but only "synchronization failed."

[yaroslav (staff)]

Hi,
Sorry for the misunderstanding. My message was a simple "don't mess up with replication". The only scenario you need to mark as synchronized
-mutual not synchronized
-the only node that is up is not synchronized.

What you did looks like breaking the replication link and attempting to override the priority-driven replication drop. Establish the connection and replication starts on its own.