Force failover

Submitted by jafinn on Sun, 11/17/2019 - 08:46

I've set up a failover group according to this https://www.virtualmin.com/documentation/cloudmin/vm/failover but the option to manually fail over seems to be missing from the web GUI? The manual states that "you can force a failover at System Operations -> Force Failover" but there is no such option either under the hosts or the VMs. I've tried both Authentic theme and the old Webmin theme to check if it's a theme issue.

I've set up the failover group with two host machines and a single VM to be manually failed over. The VM is running on a glusterfs brick mounted on both machines.

Status: 
Active
Files: 
Image icon disk_image_dir.png
Image icon failover_group.png

Comments

Submitted by jafinn on Sun, 11/17/2019 - 10:14

Some screenshots from the setup. /srv/glusterfs/pfsense is mounted on both hosts and the contents sync as expected.

From the cli I get this

# cloudmin failover-system --host test.box

No suitable failover targets were found :

debian.box : Failover is not enabled for this system in the group
hp.home : Failover is not enabled for this system in the group
debian.box : Cannot failover to current host
hp.home : Not a suitable failover target, or missing shared storage

So apparently I'm missing something important.

The second host has the same disk image directory set but are unaware of any KVM guests hosted as they are all added from the master host.

Submitted by JamieCameron on Sun, 11/17/2019 - 20:24

Are the VMs on the current system on a filesystem that is shared with the new host? If so, I'd be interested in seeing the output from the mount command on both systems.

Submitted by jafinn on Mon, 11/18/2019 - 01:18

I set up a NFS share as well since that's what you have in the manual but get the exact same error message.

Mount command for both systems included below, I've removed everything but the relevant lines for readability, let me know if you want the full output.

Machine 1

debian.box $ mount
localhost:/brick1 on /srv/glusterfs/pfsense type fuse.glusterfs (rw,relatime,user_id=0,group_id=0,default_permissions,allow_other,max_read=131072,_netdev)
192.168.1.100:/media/media on /srv/temp type nfs4 (rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=192.168.1.100,local_lock=none,addr=192.168.1.100,_netdev)

Machine 2

hp.home $ mount
localhost:/brick1 on /srv/glusterfs/pfsense type fuse.glusterfs (rw,relatime,user_id=0,group_id=0,default_permissions,allow_other,max_read=131072,_netdev)
192.168.1.100:/media/media on /media type nfs4 (rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=192.168.1.98,local_lock=none,addr=192.168.1.100,_netdev) # cloudmin failover-system --host temp.box --dest hp.home --force
System is not part of any usable failover groups

The img is available from both machines

$ ls /srv/temp/temp.box*
/srv/temp/temp.box.console   /srv/temp/temp.box.img
/srv/temp/temp.box-eth0.sh   /srv/temp/temp.box.monitor
/srv/temp/temp.box-eth0.tap

I should also probably mention that one of the machines is also a host of that could be an issue