Oh, and after you added the votes, delete the PVE VM and delete it with the
pvecm delnode oldvmnode-hostname
from the cluster
On 09/24/2015 05:08 PM, Thomas Lamprecht wrote:
Adapt the corosync.conf file in the /etc/pve folder there, set
quorum_votes of one of the node entries from 1 to 2.
This achieves the same as your current setup but with much less
overhead as no VM, which acts as PVE node, is needed.
*NOTE*: for real HA three real nodes are needed.
If you really want this behavior use the way described above, it's the
better way.
On 09/24/2015 04:24 PM, Gilberto Nunes wrote:
Hi guys
I have a three nodes cluster with PVE 4 and HA enable.
Just one VM with 8GB memory and 1,5 TB of hard disk.
If the VM has 2 GB of memory and 32 GB of hard disk, HA works fine.
But now, with more memory and hard disk, the HA freeze the VM:
ha-manager status
quorum OK
master proxmox02 (active, Thu Sep 24 11:22:10 2015)
lrm proxmox01 (old timestamp - dead?, Thu Sep 24 11:14:03 2015)
lrm proxmox02 (active, Thu Sep 24 11:22:15 2015)
lrm proxmox03 (active, Thu Sep 24 11:22:13 2015)
service vm:120 (proxmox01, freeze)
The VM reside in NFS storage...
Any idea?
--
Gilberto Ferreira
+55 (47) 9676-7530
Skype: gilberto.nunes36
_______________________________________________
pve-user mailing list
pve-user@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
_______________________________________________
pve-user mailing list
pve-user@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
_______________________________________________
pve-user mailing list
pve-user@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
