Qlik Replicate upgrade on a Windows Cluster from 2021.5 to 2021.11 and 2022.5

[Steve_Nguyen]

This article provides the steps needed to successfully upgrade Qlik Replicate installed on a Windows Cluster.

Environment:

• Qlik Replicate 2021.5 upgrade to 2021.11
• Qlik Replicate upgrade 2021.11 to 2022.5 or 2022.11

Resolution

1. Stop all tasks. Note that some tasks could take up to 20 minutes or more to stop. Monitor the process before you continue.
2. Assuming we are on Active Node A: from the Cluster Manager, take the Attunity Replicate Server and Attunity Replicate UI server Offline
3. Take a backup of all your task definitions on Node A:Qlik Replicate Command Line Console as Administrator
4. Run:

   repctl -d "YOUR_DATA_DIRECTORY_PATH" exportrepository

   Example: ~Program Files\Attunity\Replicate\bin> repctl -d "S:\Programs\Qlik\Replicate\data" exportrepository
   This example writes the backup to the Replication_Definition.json file.
    
5. Back up your data directory by copying it to a folder location outside the Qlik Replicate directories. You can skip copying the \logs subfolder.
6. On Node A: verify the installed version of Attunity Replicate. This can be done by reviewing the program installed through the Windows Add/Remove programs feature. Note: attempting to “upgrade” with an identical version will lead to Attunity Replicate being uninstalled.
7. Verify what account the Attunity (or Qlik) Windows Services are run with. Note the account credentials (username and password). After the upgrade, the account may have been switched back to Local System and the account credentials will need to be entered again. 
8. On the active node: perform the upgrade by running a newer version of the Qlik Replicate executable.
   • Right-click on the executable and run as administrator.
   • Follow the on-screen instructions.
9. On Node A:
   • Verify that the Qlik Replicate Server and Qlik Replicate UI Server are stopped in the Cluster Manager
   • Verify that the Qlik Replicate Server and Qlik Replicate UI Server are stopped in the Windows Services 
10. Move the Cluster service (this is the role that manages your share drive/Qlik service/and client access point) to another node (Node B in our example). This is going to move the shared drive and service to the other node (Node B).
11. On Node B: run the same upgrade as previously done on Node A, repeating steps 8 to 9.
12. Repeat the upgrade steps on any additional nodes. 
13. Verify the account running the Attunity Replicate Services on all nodes, changing the account back to the account which was used before the upgrade (see step 7).
14. From the Windows Cluster, bring the Attunity Replicate Server and Attunity Replicate UI Server back online.
15. Resume the tasks. Note: resume tasks at a rate of five (maximum) at a time until all have been launched.

[MoeE]

Hi, 

I have a question. If the Shared Disk is in its own role, is that okay? Will it still need to be moved to each node for the upgrade? 

Regards,

Mohammed 

[Dana_Baldwin]

Hi @MoeE 

Yes, each time you run the installer to update the software, the data directory has to be active on that node.

Thanks,

Dana

[MoeE]

Hi @Dana_Baldwin,

I have another question posed by a client. Are there any issues with upgrading the passive node first while allowing the active node to remain functioning?

Our client has the requirement of minimal downtime.

Regards,

Mohammed

[Dana_Baldwin]

Hi @MoeE 

When the upgrade runs it needs the data directory to be available on the node, so it has to be the active node. Changes to the data folder might be made during an upgrade.

Thanks,

Dana