Node shutdown while holding partner aggregate

[Domi]

We've ran into the HW-Issue where one Node shuts down because the SP Heartbeat was missed.
Node-A goes offline and comes back online.
Node-B has taken over successfully.
However, the Giveback of one SFO Aggregate was vetoed because of CIFS Locks.
We need to force the Giveback but have to wait a couple of weeks until a maint. window is approved since critical Applications work on these CIFS Shares.
What happens now if Node-B goes offline while holding the SFO-Aggregate of Node-A?
Is the Giveback forced or is the Aggregate going offline?

Thank you in advance!

[chenguanghui]

hello：

  Aggregation toes can lead to risks of data access interruption, application exceptions, and write operations. If no aggregation has been returned yet, add the -override-vetoes parameter to the returned command and set it to true.

Best regards.

[Tas]

Not if you are running mission critical apps.  I have the same issue and in past was advised to override vetoes.  As you can very well guess, SMB is okay if a critical app is an Office App, but all other client/server, or db type apps will barf.  I can't answer honestly to what will happen, I think you need to open a support call to get the real answer.  You can also try Discord, there are a few A-team members there that may be able to respond.

[Tas]

Tell me it wasn't after an update to 9.14.1P12?

[Domi]

Hi,
no, there was no Update we are running on 9.13.1P13.
The takeover happened because of an unexpected shutdown due to a HW-Issue.

[Gmox]

Hello,

What happens now if Node-B goes offline while holding the SFO-Aggregate of Node-A?

Unfortunately, if the node goes down during the takeover, all SFO aggregates will be down too.