Upgrade from standalone to HA
This feature is available only from ThingPark Enterprise 6.0.
If you want to upgrade an existing ThingPark Enterprise from Standalone to High Availability (HA), proceed as follows:
First, be sure that the two new TPE nodes are already installed with the same TPE ISO as the existing node.
Connect to Cockpit GUI.
Go to the "TPE Services" menu.
Under the TPE cluster operations menu, click on Upgrade to High Availability.
Fill in the "IP address" and the "Support password" of the three nodes (the first node IP address must be already filled with the current node IP address).
Click on Next.
Set the following fields:
List of NTP servers: List of hostnames or IP addresses (comma-separated values) of NTP servers.
Installation repository: URL of the software repository:
Leave it as it is if you want to do the upgrade from the Internet.
Or set the local repository URL if you want to do the upgrade in local mode. See Local repository installation to start the local repository and get the local repository URL.
Proxy URL: URL of the proxy server used for access to the Installation repository (used only for the upgrade from the Internet)
Click on Next.
The next step is cluster verification:
- Ensure the nodes' connectivity.
- Ensure the nodes' requirements.
If no error is detected, click on Upgrade.
At the end of the upgrade to HA, if successful, you are prompted to go to "TPE Configuration".
Click on Configure ThingPark Enterprise.
Under the "TPE Configuration" menu, fill in the "Secondary network server address" field. Usually set with the server IP address of the second node.
You can also configure the Virtual IP settings if you want to use a floating IP address to support a cluster node failover.
Click on Save & Apply.
Once done, go to the "TPE Services" menu and check that all services are well in running state and that the three TPE nodes are in ready state.
If, for some reason, the Upgrade to HA fails and the Upgrade to High Availability is no more available under the TPE cluster operations menu, in such case use the Redeploy cluster under the same menu.