cannot migrate HA enabled VM

[svacaroaia]

I had a crash on one of my servers - all VMs have been moved by HA to other servers
However, they all have been moved to only 2 servers instead of being distributed evenly
so I ended up with 2 extremely busy servers and 3 completely empty

Question:
what is the algorithm / criteria used by HA to redistribute VMs ?
Is there anything that I should do to make sure the least busy server gets the VMs?

Related to the above I cannot do an online migration of some HA enabled VM.
The error message is :
TASK ERROR: command 'clusvcadm -M failed: exit code 255
If I remove the VM from HA, migration works fine

Strangely enough, only some VMs have this issue - others are working fine

Any help would be greatly appreciated

clustat
Cluster Status for bs01a-cluster01 @ Tue Jul 30 09:02:49 2013
Member Status: Quorate


Member Name ID Status
------ ---- ---- ------
bs01a-hst01 1 Online, rgmanager
bs01a-hst02 2 Online, rgmanager
bs01a-hst09 3 Online, rgmanager
bs01a-hst10 4 Online, rgmanager
bs01a-hst11 5 Online, Local, rgmanager
bs01a-hst12 6 Online, rgmanager


pveversion -v
pve-manager: 3.0-20 (pve-manager/3.0/0428106c)
running kernel: 2.6.32-20-pve
proxmox-ve-2.6.32: 3.0-100
pve-kernel-2.6.32-20-pve: 2.6.32-100
pve-kernel-2.6.32-19-pve: 2.6.32-96
pve-kernel-2.6.32-18-pve: 2.6.32-88
lvm2: 2.02.95-pve3
clvm: 2.02.95-pve3
corosync-pve: 1.4.5-1
openais-pve: 1.1.4-3
libqb0: 0.11.1-2
redhat-cluster-pve: 3.2.0-2
resource-agents-pve: 3.9.2-4
fence-agents-pve: 4.0.0-1
pve-cluster: 3.0-4
qemu-server: 3.0-15
pve-firmware: 1.0-22
libpve-common-perl: 3.0-4
libpve-access-control: 3.0-4
libpve-storage-perl: 3.0-6
vncterm: 1.1-3
vzctl: 4.0-1pve3
vzprocps: 2.0.11-2
vzquota: 3.1-2
pve-qemu-kvm: 1.4-12
ksm-control-daemon: 1.1-1

 

[dietmar]

Question:
what is the algorithm / criteria used by HA to redistribute VMs ?
Is there anything that I should do to make sure the least busy server gets the VMs?

You can try to use different failover domains (see 'man rgmanager'). A single unordered failover domain should distribute VMs randomly.

 

[davlaw]

I had something similar, but mine was without HA. I had a snapshot attached to the VM, snapshot was stored local, VM was on shared storage. It would fail. In my case the snapshot did not exist at all on the local share, and I removed it (via GUI, had to put a dummy file in place so the GUI would have something to delete) and the migration then went fine after that.

Hope some of this made sense