Hi All,

I'm running a critical application where we are using IBM RDQM HA/DR solution. 

This means that we have 6 nodes, where 3 of them compose the Main HA group and the other 3 compose the DR HA group.

 I have a case where I need to replace the DR nodes for a new set of machines. This means that I need to recreate the DR HA group. 

I'm testing this and once I configure the new DR HA group, looks like that the main HA group (using rdqmstatus -m TEST -a ) its still looking for the old set of. 

e.g. 

getting this:

DR remote IP address list:              10.1.111.111,10.1.111.112,10.1.111.113

expecting this

DR remote IP address list:              10.1.111.114,10.1.111.115,10.1.111.116

So without recreating the Queue Manager is it possible to reconfigure a new DR HA group with new set of IPs?

From the documentation seems possible:https://www.ibm.com/docs/en/ibm-mq/9.4.0?topic=availability-replacing-failed-node-in-drha-configuration, But I'm seeing where and how we can change the replication on main HA group to start pointing to the new IP's of the new DR HA group.

If I remember well the documentation tells you how to change out one of the three servers. And there is no changing IP's . You'll need to swap them out so that the new server has the same IP as the old one...

Hope it helps

Hi Francois,

Thanks for replying. 

From your comment I'm understanding that in case of a disaster/recovery scenario where machines are not possible to recover with the same IP's we need to start everything from zero even if one HA group is running fine? 

I mean the DR replication from the RDQM HA/DR is not reconfigurable at all?

If you can't keep the same  IPs / dns names, you have to start back from zero. You can backup  the queue manager from the still running part of the cluster (you'll need to stop it first) and then restore it unto the new RDQM cluster.