TOPIC:

Hi,

I've finally got HA-Lizard 2.1.0 installed correctly on my 2-node XenServer 7.0 pool. I've transferred my VMs to the new HA iSCSI volume, I've set global_vm_ha to 0 and have specified which of my VMs I'd like to have high availability.

What is the best way to emulate a scenario where my primary server fails?
Could I just pull the ethernet cables at the back?
What would be the procedure to get it back online afterwards?

Thanks

Running through test scenarios is best done on a test pool.

Attached are some of the basic scenarios we test and the expected results.

Great test data, thank you.

If I lose the master temporarily, and the slave becomes the new master and starts hosting the VMs, what happens when I eventually get the old master back online? Will it take on the role of slave automatically? Is there a best practice for this scenario?

The old master will rejoin the pool as a slave in this scenario.

Hi, how did you solve your problem generated by converting local storage to iSCSI ?
Seems i have the same errors like you.
I also found that solution to manually forget the local storage and then appear that HA-Lizard installed correctly and now the first sync is running.

The only problem i have is "No LUN's were found on 10.10.10.3"

Thanks

To reasing LUN use

tgt-admin --dump | grep -o "backing-store \(.*\)" | sed -e "s/^backing-store \(.*\)$/\1/"

and nothing on replay


tgt-admin --update ALL
service tgtd restart

to see if running
tgt-admin --dump | grep -o "backing-store \(.*\)" | sed -e "s/^backing-store \(.*\)$/\1/"

will echo

/dev/drbd1