*SOLVED* VE monitoring [message #12972] |
Wed, 16 May 2007 08:41 |
pege
Messages: 8 Registered: May 2007
|
Junior Member |
|
|
Hello everybody!
I'm building a server system with two identical hardware nodes with identical VE:s inside and a hearbeat between them. What I'd like is for the active machine to be able to tell if all the VE's are running ok, and if not, using heartbeat switch to the other server. How would I do this? Is there some utility for monitoring the VE's and their function? Can a VE for example kernel panic without anyone noticing? Will 'vzlist -a' show it running or stopped?
So many questions...
Thanks for any help on this.
[Updated on: Fri, 18 May 2007 07:29] by Moderator Report message to a moderator
|
|
|
Re: VE monitoring [message #12977 is a reply to message #12972] |
Wed, 16 May 2007 11:52 |
Vasily Tarasov
Messages: 1345 Registered: January 2006
|
Senior Member |
|
|
Hello,
There is an article about using heartbeat and OpenVZ together http://wiki.openvz.org/HA_cluster_with_DRBD_and_Heartbeat
Quote: | Can a VE for example kernel panic without anyone noticing?
Will 'vzlist -a' show it running or stopped?
|
As you now in OpenVZ one kernel is used for all VEs. So if kernel panics, you'll info in /var/log/message, dmesg and console as usual. vzlist's output will depend on whether all processes die in VE or not...
Vasily.
|
|
|