HA VM's

[John Wick]

Hi,

Today i want to test HA on VM's guest. So we have 3 node. All VM's test already include under group of HA. We pull out one of the node. But from HA, we saw state on every VM's is freeze. Is it normal? Suppose it goes to other node, am i right?

Thanks.

 

[dietmar]

Hard to say without any information about your HA configuration. How did you pull out the node (shutdown? or reboot/stop ?)

 

[dcsapak]

We pull out one of the node

how did you pull out the node?

 

[John Wick]

i just reboot only. Do i need to shutdown it?

 

[czechsys]

i just reboot only. Do i need to shutdown it?

You need hard poweroff, not reboot/shutdown.

 

[John Wick]

You need hard poweroff, not reboot/shutdown.

We have test HA. Is it normal console VM's disconnect while VM's transfer to other node? It's take a few minute before it's start on other node.

 

[elasc]

I would like to follow-up this thread. After setting up my 3rd node, I also configured a vm for ha and powered the node off with poweroff command on the console. Why is it needed to poweroff the node hard? Can't understand that...

 

[dietmar]

Why is it needed to poweroff the node hard? Can't understand that...

You only need to poweroff of the node to test what happens on power loss.

 

[elasc]

@dietmar got that. So there is no way to migrate vms online to other nodes automatically with one command (which will do the migration and poweroff the node afterwards)?

 

[dcsapak]

So there is no way to migrate vms online to other nodes automatically with one command (which will do the migration and poweroff the node afterwards)?

see the webinterface node -> bulk migrate
or the api call '/nodes/<NODENAME>/migrateall'