Hello!
I have 3 servers, on every server I installed Proxmox 5.2-9 and joined them to one cluster (with HA, softdog).
As shared storage I'm using CEPH (on 10GbE NICs).
I'm doing this scenario to test if HA is working:
1) Install on first node two VMs (one is Windows 7 and second is CentOS 7).
2) After both VM is fully booted, migrate them to node 2 or 3 (just for sure, to test that all is ok with network etc.)
3) Back this two VMs to node1
4) Kill the node1 by cutting the power (by IPMI).
I opened a view to the cluster from node3, so I can see what happens next.
After few seconds node1 is marked with red dot. Now HA start to migrate both VMs to other node (node2) and power up them. Also in HA tab both VMs are marked as "started".
But when I go to Console on VM I can't connect to VM by VNC, also I see on Summary tab that machine is in running state, but don't using any CPU shares (showing all of time 0.00%). State of machine is like in "freeze" mode, but he never start.
When I power up server node1 and he back to the cluster both VM (running currently on node2)
immediately start working. I can connect to the Console where I see that machine is normaly booting.
I want to know why this happen? Why when I kill node1 and both VMs are migrated to node2 where stay "hold in started state" I can't use it - and when I restore the power of node1, after proxmox boot, imidietlny VM running on node2 start working.
Thanks of any advice!
I have 3 servers, on every server I installed Proxmox 5.2-9 and joined them to one cluster (with HA, softdog).
As shared storage I'm using CEPH (on 10GbE NICs).
I'm doing this scenario to test if HA is working:
1) Install on first node two VMs (one is Windows 7 and second is CentOS 7).
2) After both VM is fully booted, migrate them to node 2 or 3 (just for sure, to test that all is ok with network etc.)
3) Back this two VMs to node1
4) Kill the node1 by cutting the power (by IPMI).
I opened a view to the cluster from node3, so I can see what happens next.
After few seconds node1 is marked with red dot. Now HA start to migrate both VMs to other node (node2) and power up them. Also in HA tab both VMs are marked as "started".
But when I go to Console on VM I can't connect to VM by VNC, also I see on Summary tab that machine is in running state, but don't using any CPU shares (showing all of time 0.00%). State of machine is like in "freeze" mode, but he never start.
When I power up server node1 and he back to the cluster both VM (running currently on node2)
immediately start working. I can connect to the Console where I see that machine is normaly booting.
I want to know why this happen? Why when I kill node1 and both VMs are migrated to node2 where stay "hold in started state" I can't use it - and when I restore the power of node1, after proxmox boot, imidietlny VM running on node2 start working.
Thanks of any advice!