Hi Ashlin,

In a RDQM HA group, you can perform a managed failover for a queue manager by changing its preferred node.

See https://www.ibm.com/docs/en/ibm-mq/9.3?topic=availability-setting-preferred-location-rdqm

You can set a single preferred node, or you can set a first and a second preference. If you don't set a preferred node, then Pacemaker can select any one of the nodes. If a preferred node is set and that node goes down, such as for maintenance, then queue managers will fail over to another node (e.g. their second preference). Once the original node is back up then the queue managers will fail back if it is their first preference. If you do not want the additional fail over for queue managers to return to that node then you can update or clear their preferred node setting so that the queue managers remain running on their current node instead.

Kind regards,

Jamie

Thanks a lot Jamie, it's helpful! 

Hi Ashlin,

In a RDQM HA group, you can perform a managed failover for a queue manager by changing its preferred node.

See https://www.ibm.com/docs/en/ibm-mq/9.3?topic=availability-setting-preferred-location-rdqm

You can set a single preferred node, or you can set a first and a second preference. If you don't set a preferred node, then Pacemaker can select any one of the nodes. If a preferred node is set and that node goes down, such as for maintenance, then queue managers will fail over to another node (e.g. their second preference). Once the original node is back up then the queue managers will fail back if it is their first preference. If you do not want the additional fail over for queue managers to return to that node then you can update or clear their preferred node setting so that the queue managers remain running on their current node instead.

Kind regards,

Jamie