StarWind Software

Have a setup with 2 nodes of latest V8 build. During storage migration of VMs from local storage to shared storage over SMB to failover cluster backed by StarWind, both nodes lose connection and service hangs during restart.

iSCSI/Mirror/Synchronization over 10Gbe
Synchronization/SMB over 10Gbe - also Synchronization
Management 1Gbe for heartbeat

StarWind V8 on Windows Server 2012 R2 with SoFS clustering
MPIO setup
iSCSI points to local, then iSCSI, and then SMB for failover only
SMB traffic from Hyper-V restricted to Sync/SMB 10Gbe

Logs/dumps can be downloaded from here [ ... ]

Hi!

Sorry to hear that you are in trouble.

both nodes lose connection and service hangs during restart.

May I ask you to clarify what exactly have you tried to restart?

iSCSI/Mirror/Synchronization over 10Gbe

So, you are running iSCSI initiator and synchronization traffics on the same links? If so, then you should know that this is bad idea.

Logs/dumps can be downloaded from here [ ... ]
There is nothing there.

Even with 2 different 10Gbe nics?

I will re-upload the file and post link unless you have a better place to upload.

The service I restart is the StatWind Virtual SAN for vSphere.

Even with 2 different 10Gbe nics?

Yes. SyncChannel is really important, so you should keep it separated from other traffic.

I will re-upload the file and post link unless you have a better place to upload.

Well, you can try dropbox, box, gDrive, etc.

The service I restart is the StatWind Virtual SAN for vSphere.

So, just to confirm, the client machine is loosing the connection to all StarWind targets after you are restarting the service on one of the node?