Hi,

I am thinking of the following scenario (without a real stonith device)..

There is node A with resource r1 and B without resources and they are in a
split-brain condition.
Lets say the quorum server decides to make node B the DC.
So A tries to stop resource r1, but this will not work for any reason.

Because of the resource stop failure or the lost quorum node A decides to
kill itself and reboot.. (How ever this is done with V2? A node really
should have the possibility to kill itself..)

So lets say A comes back after reboot and the network works again.
B is still trying to stonith A because it doesn't know of that reboot.
Is it possible that A sends its uptime to B, so B knows ok everything
clear again?

If not, wouldn't that be a good idea?

Cheers,
Holger

_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to