Uninstalling RDQM (replicated data queue managers) and upgrading

How to uninstall RDQM in preparation for upgrading IBM® MQ and RDQM.

Depending on which version of IBM MQ is installed, you might need to remove maintenance before you uninstall the base packages:

If you are uninstalling a version of IBM MQ at IBM MQ 9.4.0 or later, you do not need to remove maintenance before you uninstall IBM MQ.
If you are uninstalling a version of IBM MQ before IBM MQ 9.4.0, you must remove any maintenance that is applied to IBM MQ before you can uninstall. The procedure for removing maintenance changed at IBM MQ 9.4.0. Therefore, you must use the procedure that is detailed in earlier versions of the product documentation to remove the maintenance.

Important: You must stop all IBM MQ queue managers, other objects, and applications, before you begin the process to uninstall or modify IBM MQ.

Note: If you are upgrading the OS RHEL version, you need to follow a different upgrade procedure. See Applying OS updates with RDQM for upgrading RHEL within a version, or Migrating an RDQM configuration from RHEL 8 to RHEL 9 for updating versions.

This topic describes upgrading RDQM between versions. To apply maintenance level updates to RDQM, see Applying maintenance level updates for RDQM. There are different procedures for uninstalling and then upgrading a high availability (HA) configuration, a disaster recovery (DR) configuration, or a combined DR/HA configuration.

For HA configurations, complete steps on each node in the HA group in turn. Processing can continue on other nodes while this is in progress.

For all configurations, if the upgrade is to a level of IBM MQ that is at a higher command level then, after a queue manager has been started at the higher level, it cannot run on a node that has not yet been upgraded. You should plan the sequence of upgrades accordingly.

Important: The following commands are intended as an example of those that would be issued on a system with a single IBM MQ installation. On systems with multiple IBM MQ installations, or where there are other packages installed whose name includes either "drbd" or "linbit", the commands must be updated to ensure that only the packages associated with this IBM MQ installation are removed.

For details of how to uninstall IBM MQ on a system with multiple MQ installations please see Uninstalling or modifying IBM MQ on Linux using rpm.

RDQM HA

Complete the following steps to uninstall and update an RDQM HA configuration:

Suspend the HA group on the node, by entering the following command:
```
rdqmadm -s
```
Log in as root or switch to superuser using the su command.
Uninstall IBM MQ (this step also uninstalls RDQM):
```
rpm -qa | grep MQSeries | xargs dnf -y remove
```

Uninstall Pacemaker:

rpm -qa | grep linbit | xargs dnf -y remove

Uninstall DRBD:

rpm -qa | grep drbd | xargs dnf -y remove

Verify that the DRBD kernel was successfully unloaded:
```
lsmod | grep drbd
```
If either the drbd or drbd_transport_tcp kernel modules are still loaded, they can be unloaded with the following commands:
```
modprobe -r drbd_transport_tcp
modprobe -r drbd
```
If the unload fails for any reason, reboot the node.

Uninstall the tlshd service:

rpm -qa | grep ktls-utils | xargs dnf -y remove

Install the new level of IBM MQ and dependent software, see Installing RDQM (replicated data queue managers).
Resume the HA group on the node by entering the following command:
```
rdqmadm -r
```
You can now proceed to the next node in the group.

RDQM DR

Complete the following steps to uninstall and update an RDQM DR configuration:

Upgrade the DR secondary node:
1. Log in as root or switch to superuser using the su command.
2. Uninstall IBM MQ (this step also uninstalls RDQM):
```
rpm -qa | grep MQSeries | xargs dnf -y remove
```
3. Uninstall Pacemaker:
```
rpm -qa | grep linbit | xargs dnf -y remove
```
4. Uninstall DRBD:
```
rpm -qa | grep drbd | xargs dnf -y remove
```
5. Verify that the DRBD kernel was successfully unloaded:
```
lsmod | grep drbd
```
  If either the drbd or drbd_transport_tcp kernel modules are still loaded, they can be unloaded with the following commands:
```
modprobe -r drbd_transport_tcp
modprobe -r drbd
```
  If the unload fails for any reason, reboot the node.
6. Uninstall the tlshd service:
```
rpm -qa | grep ktls-utils | xargs dnf -y remove
```
7. Install the new levels of IBM MQ and RDQM, see Installing RDQM (replicated data queue managers).
On the DR primary node, do one of the following steps:
- End the DR queue managers, or
- Perform a managed failover of the DR queue managers to the DR secondary node.
Upgrade the DR primary node:
1. Log in as root or switch to superuser using the su command.
2. Uninstall IBM MQ (this step also uninstalls RDQM):
```
rpm -qa | grep MQSeries | xargs dnf -y remove
```
3. Uninstall Pacemaker:
```
rpm -qa | grep linbit | xargs dnf -y remove
```
4. Uninstall DRBD:
```
rpm -qa | grep drbd | xargs dnf -y remove
```
5. Verify that the DRBD kernel was successfully unloaded:
```
lsmod | grep drbd
```
  If either the drbd or drbd_transport_tcp kernel modules are still loaded, they can be unloaded with the following commands:
```
modprobe -r drbd_transport_tcp
modprobe -r drbd
```
  If the unload fails for any reason, reboot the node.
6. Install the new levels of IBM MQ and RDQM, see Installing RDQM (replicated data queue managers).
On the DR primary node, do one of the following steps:
- Start the DR queue managers (if you previously ended them), or
- Perform a managed failover of the DR queue managers back to the DR primary node.

RDQM HA/DR

Complete the following steps to uninstall and update an RDQM HA/DR configuration:

Upgrade the HA group on your recovery site (presuming that the DR/HA RDQMs are running on the main site). Complete the following steps on each node in the group in turn.
1. Log in as root or switch to superuser using the su command.
2. Suspend the HA group on the node, by entering the following command:
```
rdqmadm -s
```
3. Uninstall IBM MQ (this step also uninstalls RDQM):
```
rpm -qa | grep MQSeries | xargs dnf -y remove
```
4. Uninstall Pacemaker:
```
rpm -qa | grep linbit | xargs dnf -y remove
```
5. Uninstall DRBD:
```
rpm -qa | grep drbd | xargs dnf -y remove
```
6. Verify that the DRBD kernel was successfully unloaded:
```
lsmod | grep drbd
```
  If either the drbd or drbd_transport_tcp kernel modules are still loaded, they can be unloaded with the following commands:
```
modprobe -r drbd_transport_tcp
modprobe -r drbd
```
  If the unload fails for any reason, reboot the node.
7. Uninstall the tlshd service:
```
rpm -qa | grep ktls-utils | xargs dnf -y remove
```
8. Install the new level of IBM MQ and dependent software, see Installing RDQM (replicated data queue managers).
9. Resume the HA group on the node by entering the following command:
```
rdqmadm -r
```
  You can now proceed to the next node in the group.
On the HA group at the main site, either stop your queue managers, or perform a managed failover to the HA group that you have just upgraded on the recovery site.
Upgrade the HA group on your main site. Complete the following steps on each node in the group in turn.
1. Log in as root or switch to superuser using the su command.
2. Suspend the HA group on the node, by entering the following command:
```
rdqmadm -s
```
3. Uninstall IBM MQ (this step also uninstalls RDQM):
```
rpm -qa | grep MQSeries | xargs dnf -y remove
```
4. Uninstall Pacemaker:
```
rpm -qa | grep linbit | xargs dnf -y remove
```
5. Uninstall DRBD:
```
rpm -qa | grep drbd | xargs dnf -y remove
```
6. Verify that the DRBD kernel was successfully unloaded:
```
lsmod | grep drbd
```
  If either the drbd or drbd_transport_tcp kernel modules are still loaded, they can be unloaded with the following commands:
```
modprobe -r drbd_transport_tcp
modprobe -r drbd
```
  If the unload fails for any reason, reboot the node.
7. Uninstall the tlshd service:
```
rpm -qa | grep ktls-utils | xargs dnf -y remove
```
8. Install the new level of IBM MQ and dependent software, see Installing RDQM (replicated data queue managers).
9. Resume the HA group on the node by entering the following command:
```
rdqmadm -r
```
  You can now proceed to the next node in the group.
You can now either start your queue managers (if you previously stopped them) or fail them back over to the main site from the recovery site.