Re: Partitioned cluster question (reboot during lost quorum)

In article <44453DE5.2759B34E@xxxxxxxxxxxx>, JF Mezei
<jfmezei.spamnot@xxxxxxxxxxxx> writes:

Say you have a 3 node cluster, node1, node2 node3

Each node has 1 vote. You need 2 votes to maintain quorum.

You unplug the ethernet from node3. node1 and node2 still have quorum
and happily chug along. Node3 realises it has lost quorum and freezes.

What happens if at this point, you reboot node3 ?

It will reboot, then wait to form or join an OpenVMS cluster.

Node3 would not see 1 or 2 and thing it was the first node rebooting
after say a power failure. Right ?

If it doesn't see 1 or 2, it assumes EITHER that it is the first node
rebooting after say a power failure OR that there is a connectivity
problem.

Wouldn't NODE3 then form its instance of the cluster with its own vote
and a quorum of 1 ?

No, since it will have EXPECTED_VOTES=3. That's the purpose of
EXPECTED_VOTES.

Or is there anything which would prevent node3 from rebooting ?

No, it will reboot, but hang when waiting to form or join a cluster.

Quorum is half + 1 of VOTES or EXPECTED_VOTES, whichever is GREATER.
Whenever you add a voting node to the cluster, EXPECTED_VOTES goes up,
and stays up. You can reduce it manually, if you plan to take a node
out of the cluster (the "remove node" option in SHUTDOWN.COM also does
this), but if you do nothing, it will always reflect the number of votes
during the last instance of the cluster.

Once node3 has rebooted, thinking it was alone in the cluster,

It won't think that since EXPECTED_VOTES was too high.

happens when you plug the ethernet back in and all of a sudden, node3
sees nodes 1 and 2 ? Will nodes 1 and 2 succeed in convinding node3 to
commit suicide ? Will node3 convince nodes1 and 2 to commit suicide
(since it is a more recent incarnation of the cluster) ?

Or will what is essentially a partitioned cluster continue to exist as 2
separate cluster instances ?

None of the above.

If you want to boot a machine which has been part of a larger cluster,
you'll have to do a minimum boot and reset VAXCLUSTER to 0, or reduce
EXPECTED_VOTES, or increase VOTES.

.

Relevant Pages