Home > Lab 04. Admin Cluster > Task 05 - Upgrade Cluster Software(Optional)

Task 05 - Upgrade Cluster Software(Optional)

Depending on the update status at the time of the lab, the update may not be applied or the state of the currently configured HA cluster may affect unexpected test scenarios. If you need to troubleshoot an unexpected impact, it will be outside the scope of this workshop and will affect the progress of the Lab 06. Change to Cost Optimized Scenario that will be conducted later. We recommend that you do it for the last time after all labs are over.

In this lab we will show you how to update SUSE Linux on both cluster nodes.

Update Cluster Software

Stop or Maintenance Cluster
- If a package update requires a reboot, stop the cluster stack on the secondary node before starting the software update.
```
sudo su -
crm cluster stop
```
- If not needed, switch to maintenance mode
```
crm node maintenance {NODE_NAME}
```
Install package updates :
```
sudo su -
zypper up
```
Show available patches :
```
sudo su -
zypper lp
```
Show available updates :
```
sudo su -
zypper lu
```
Start or Off Maintenance Cluster
- Start the cluster on each node or restart the server.
```
sudo su -
crm cluster start

or

sudo su -
reboot
```
- Off Maintenance Cluster
```
crm node maintenance {NODE_NAME} off
```
Lab 03. Execute Task 01. Validation of Cluster to check the cluster status through CLI.
Lab 02. Task 03. Execute HAWK (High Availability Web Konsole) to check the cluster status through the web.

Performing Cluster-wide Rolling Upgrade

If you are performing rolling upgrade of the entire cluster, you need to take one node offline to proceed with the upgrade, then take back and repeat the upgrade procedure on the other node.

crm cluster stop

Perform the upgrade to the desired target version of SUSE Linux Enterprise Server for SAP applications
Start the cluster stack on the node where the upgrade has been completed, allowing the node to rejoin the cluster.

crm cluster start

Before moving to the next node, check the cluster status with the command below or HAWK.

crm_mon -rfn1

Take the next node offline and repeat the procedure for that node.