Upgrading ThingWorx High Availability Clustering
Upgrading from a Single Server to a ThingWorx High Availability Cluster
If you want to upgrade your ThingWorx single-server instance to operate in a high availability environment in cluster mode after you upgrade to ThingWorx 9.x.x, see Overview of ThingWorx High Availability Clustering for configuration details.
Upgrading from an Earlier ThingWorx 9.x.x Version of ThingWorx High Availability Clustering
If you are upgrading from an earlier ThingWorx high availability cluster to a newer version of high availability clustering, refer to the Release Advisor to determine which type of upgrade you can perform.
 |
Before upgrading ThingWorx High Availability Clustering, refer to the Release Advisor to determine if a specific upgrade version is supported.
|
• Full cluster upgrade
First, determine from release advisor if your upgrade path supports a rolling cluster upgrade. If not, you will need to do a full cluster upgrade. If you’re unsure, a full cluster upgrade is always a viable option..
In general, full cluster upgrades are required if either of the following is true:
◦ There are changes to the data classes used by Apache Ignite, which can cause hydration issues when reading data from the cache.
◦ Data migration is required. Therefore, old systems may not be able to access certain data.
To perform a full cluster upgrade, do the following:
a. Stop all ThingWorx servers.
Allow the servers to shut down completely and drain their data queues.
b. Shut down all Ignite servers.
c. Once the Ignite cluster is shut down, perform the required data migration and schema changes to the persistence provider database(s). For more information, see the Upgrading ThingWorx section.
d. Restart Ignite. For more information, see Installing and Configuring Ignite for ThingWorx HA.
e. Upgrade or migrate the ThingWorx servers as appropriate. For more information, see the Upgrading ThingWorx section.
f. Restart the ThingWorx servers
• Rolling upgrade
If your upgrade path supports a rolling upgrade for your from/to versions and the above points under the Full cluster upgrade section are not true, you can perform a rolling upgrade. A rolling upgrade allows you to keep the cluster running and take only one server offline at a time.
To perform a rolling upgrade, do the following:
a. Stop one of the ThingWorx servers in the cluster, allowing it to shut down completely and drain its data queue.
b. After the ThingWorx server has been shut down, upgrade the server to ThingWorx 9.x.x. For more information, see the Upgrading ThingWorx section.
c. Restart the upgraded ThingWorx server.
d. Repeat the upgrade and restart steps for each ThingWorx server in the cluster.