[ClusterLabs] Antw: Re: OCFS2 on cLVM with node waiting for fencing timeout

Andrei Borzenkov arvidjaar at gmail.com
Tue Oct 11 06:22:41 UTC 2016


On Tue, Oct 11, 2016 at 9:18 AM, Ulrich Windl
<Ulrich.Windl at rz.uni-regensburg.de> wrote:
>
> My point is this: For a resource that can only exclusively run on one node, it's important that the other node is down before taking action. But for cLVM and OCFS2 the resources can run concurrently on each node,

Both require coordination between nodes. They can do something on
local node only after they get in agreement with other nodes. So they
must wait for other nodes to agree or for indication that other nodes
are dead.




More information about the Users mailing list