I got so many emails that how to configure the redhat cluster to migrate redhat virtual guest(KVM guests) from one cluster node to another and we have seen how to configure it in the previous post.Once the cluster configuration is done ,we are good to test the live migration.Before proceeding to the live migration,we will do some manual test by stopping the service(VM: KVM guest) in one cluster node and starting it in another.If you didn’t have the ssh keyless authentication configured for root between two cluster nodes, you will get below error messages in /var/log/messages.
Assumptions:
Virtual machine is installed in shared location and Redhat Cluster is configured with virtual guest.
1. Check the cluster status and service status.
[root@uagl1 ~]# clustat Cluster Status for uacl1 @ Thu Oct 24 22:44:00 2013 Member Status: Quorate Member Name ID Status ------ ---- ---- ------ uagl1h 1 Online, Local, rgmanager uagl2h 2 Online, rgmanager Service Name Owner (Last) State ------- ---- ----- ------ ----- vm:VMG uagl1h started [root@uagl1 ~]#
2.Check the virtual machine status.
[root@uagl1 ~]# virsh list Id Name State ---------------------------------------------------- 26 VMG running [root@uagl1 ~]#
3. Login to any of the cluster node and stop the service(vm:VMG).This will halt the virtual machine (KVM Guest)
[root@uagl1 ~]# clusvcadm -d vm:VMG Local machine disabling vm:VMG...Success [root@uagl1 ~]# virsh list Id Name State ---------------------------------------------------- [root@uagl1 ~]# clustat Cluster Status for uacl1 @ Thu Oct 24 22:45:48 2013 Member Status: Quorate Member Name ID Status ------ ---- ---- ------ uagl1h 1 Online, Local, rgmanager uagl2h 2 Online, rgmanager Service Name Owner (Last) State ------- ---- ----- ------ ----- vm:VMG (uagl1h) disabled [root@uagl1 ~]#
4.Start the service (vm:VMG) on alternative node.
[root@uagl1 ~]# clusvcadm -e vm:VMG -m uagl2h Member uagl2h trying to enable vm:VMG...Success vm:VMG is now running on uagl2h [root@uagl1 ~]# ssh uagl2h Last login: Thu Oct 24 06:55:08 2013 from uagl1 [root@uagl2 ~]# clustat Cluster Status for uacl1 @ Fri Oct 25 04:24:28 2013 Member Status: Quorate Member Name ID Status ------ ---- ---- ------ uagl1h 1 Online, rgmanager uagl2h 2 Online, Local, rgmanager Service Name Owner (Last) State ------- ---- ----- ------ ----- vm:VMG uagl2h started [root@uagl2 ~]#
5.Check the virtual machine status in the alternative node.
[root@uagl2 ~]# virsh list Id Name State ---------------------------------------------------- 3 VMG running [root@uagl2 ~]#
We are able to migrate the virtual machine from one KVM node to another node successfully with minimal downtime.If you use the live migration option, we can move the virtual machine from one node to another node without any downtime.(On the fly). Let’s see how we can do that with help of Redhat cluster.
1. Login to any of the cluster node.
[root@uagl1 ~]# clustat Cluster Status for uacl1 @ Fri Oct 25 04:24:28 2013 Member Status: Quorate Member Name ID Status ------ ---- ---- ------ uagl1h 1 Online, Local, rgmanager uagl2h 2 Online, rgmanager Service Name Owner (Last) State ------- ---- ----- ------ ----- vm:VMG uagl2h started [root@uagl1 ~]#
2.Perform the Live-migration for service vm:VMG.
[root@uagl1 ~]# clusvcadm
usage: clusvcadm [command]
Resource Group Control Commands:
-v Display version and exit
-d Disable . This stops a group
until an administrator enables it again,
the cluster loses and regains quorum, or
an administrator-defined event script
explicitly enables it again.
-e Enable
-e -F Enable according to failover
domain rules (deprecated; always the
case when using central processing)
-e -m Enable on
-r -m Relocate [to ]
Stops a group and starts it on another
cluster member.
-M -m Migrate to
(e.g. for live migration of VMs)
-q Quiet operation
-R Restart a group in place.
-s Stop . This temporarily stops
a group. After the next group or
or cluster member transition, the group
will be restarted (if possible).
-Z Freeze resource group. This prevents
transitions and status checks, and is
useful if an administrator needs to
administer part of a service without
stopping the whole service.
-U Unfreeze (thaw) resource group. Restores
a group to normal operation.
Resource Group Locking (for cluster Shutdown / Debugging):
-l Lock local resource group managers.
This prevents resource groups from
starting.
-S Show lock state
-u Unlock resource group managers.
This allows resource groups to start.
[root@uagl1 ~]#
[root@uagl1 ~]# clusvcadm -M vm:VMG -m uagl1h
Trying to migrate vm:VMG to uagl1h...Success
[root@uagl1 ~]# clustat
Cluster Status for uacl1 @ Thu Oct 24 22:47:54 2013
Member Status: Quorate
Member Name ID Status
------ ---- ---- ------
uagl1h 1 Online, Local, rgmanager
uagl2h 2 Online, rgmanager
Service Name Owner (Last) State
------- ---- ----- ------ -----
vm:VMG uagl1h started
[root@uagl1 ~]#
3.You can see the virtual machine will not be halted or rebooted during this migration.
[root@uagl1 ~]# virsh list Id Name State ---------------------------------------------------- 27 VMG running [root@uagl1 ~]#
This migration is like vmware’s vmotion where we can move the virtual machine from one node to another node without any downtime.
Hope you enjoyed the KVM guest live migration. Please leave a comment if you have any doubt on this.
Thank you for visiting UnixArena.