Lon Hohberger wrote:
On Thu, 2005-02-24 at 14:00 +0100, Ion Alberdi wrote:Quite long, I fence_ack_manual the node and i wait but nothing happens,
Ok, because what I'm trying doesn't work:
I have three nodes debian, gump,buba.
debian exports devices to gump and buba over gnbd, and finally the three nodes have the same /dev/vg1/lv1 thanks to clvmd.
I created a script which mounts this gfs logical volume on /mnt/gfs and appends the name of the node in nodes_name.txt. (and stops writting and unmounting to stop), and I put this script as a resource.
I launch rgmanager and with clusvcadm I start, stop realocate the resource, this works very well!
But when I unplug the ethernet cable of the node running the resource, the resource isn't relocated to
another node, I suppose that's what you mean by it doesnt work with CMAN>2nodes.
If you need another tester to test your fixes here am I :)
Failover will not occur until after CMAN (or gulm) says the node is dead and has been fenced. When using the kernel Service Manager (provided by CMAN), recovery is in the following order:
(1) Fencing (2) Locking (3) GFS (4) User services (e.g. rgmanager)
How long did you wait? :)
I tried rgmanager with two nodes, and a basic script (no gfs, no logical volume, no gnbd) and it works well, I'm going to try with another script (which will use, gnbd,gfs, and logical volumes) on two nodes
-- Lon
-- Linux-cluster@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/linux-cluster
