Recover from loss of sync on 2-node VSAN Hyper-V

[james643]

Server 2019 core, Hyper-V clustering, 2-node

I believe some firewall rules caused a loss of connectivity and devices got out of sync. Firewall is opened up and there are no drops occuring now.

Node1 holds MasterImg21 which the event logs on both nodes list as most recently marked fully synchronized so I want to take that as the good image and sync it to PartnerImg22 on Node2.

I can't get iSCSI to connect from the Node2 side and believe that it's because StarWindService is reporting the devices are out of sync.

How do I get Node1's MasterImg21 marked as good so that Node2 will connect and accept sync to its PartnerImg22?

[james643]

So far I've managed to get the HAImage1 / masterimg21 on server CL1 marked as SyncStatus=1, and this did allow me to make the local iSCSI connection on CL1 to masterimg21, and also allowed the remote connection from CL2 to masterimg21.

However I still can't get either CL1 or CL2 to connect iSCSI to the HAImage2's partner side file 'partnerimg22' or either to connect to the witness. I'm assuming this is an identical issue but I am not sure what to do with the sync status on partnerimg22 and am afraid to guess.

Here is what I did to get the image file on CL1 marked as ok:

Code: Select all

Import-Module StarWindX; Enable-SWXLog
#Log into the swserver
$addr="127.0.0.1"; $port=3261; $user="root"; $password="starwind"; $deviceName="HAImage1"
$server = New-SWServer $addr $port $user $password; $server.Connect()
$device = Get-Device $server -name $deviceName

#Show info about the devices
$server
$device

#Confirmed that HAImage1 here is the most recent synced one that I want, iqn.2008-08.com.starwindsoftware:cl1.domain.local-masterimg21
#Mark it as synced then re-check it for SyncStats -eq 1
$device.MarkAsSynchronized()
$device = Get-Device $server -name $deviceName
$device.SyncStatus

Any assistance on getting the partner image to allow connections to its iSCSI targets greatly appreciated.

EDIT: Partner node fix might already be happening? Event ID 780 shows:

HA Device iqn.2008-08.com.starwindsoftware:cl1.domain.local-masterimg21: Full synchronization of partner node iqn.2008-08.com.starwindsoftware:cl2.domain.local-partnerimg22 started.

Edit #2: after setting the Witness on CL1 to status Synced it still reported that it was NOT in SyncStatus = 1, but checking the event log later it said syncing. I just looked at the MS cluster console and Witness is green. Sync status on CL2 from the included script says CL2's partnerimg22 file is now at 34% so it looks like I may have gotten this solved:

Server: 192.168.1.111 - HAImage1
SyncStatus: synchronized
SyncPercent: 0 %
-----------
Server: 192.168.1.121 - HAImage1
SyncStatus: synchronizing
SyncPercent: 34 %
-----------

[yaroslav (staff)]

Hi,

Please try restarting the service on the not synchronized node. For v8.0.0.12658, it may be necessary restarting the service on the active node AND update once the sync is over.
Let me know if this advice does not help.