Question
HiveMQ 4.7 reached its end of life. What is the recent long-term support version and what is the upgrade path?
Answer
Current LTS version
HiveMQ 4.9 - Current Long-Term Support (LTS) version (End of life: 28-09-2024). For the full list of all versions and their status of support please refer to the page Download supported HiveMQ products.
Upgrade Path
Please refer to the HiveMQ Upgrade Guide section in our Documentation: https://www.hivemq.com/docs/hivemq/4.9/upgrade/introduction.html
Chose which type of upgrade is more suitable for your HiveMQ cluster. With a rolling upgrade, you add a node with the desired new version to your cluster and then shut down one cluster node that currently runs the old version, then you update the node that you shut down with the desired new version and restart the updated node. In a rolling upgrade, only part of the cluster capacity for an application is offline at a given time. Regarding the rolling upgrade, refer to the Rolling Upgrades section in our Documentation: https://www.hivemq.com/docs/hivemq/4.9/user-guide/cluster.html#rolling-upgrades
If you do not require a rolling upgrade, it means that you can cut off all clients and shut down the cluster, upgrade every single node, start nodes and then allow the clients to connect to the cluster.
For the upgrade steps please refer to the version upgrade guides: https://www.hivemq.com/docs/hivemq/4.9/upgrade/4-7-to-4-8.html#cluster-upgrade and https://www.hivemq.com/docs/hivemq/4.9/upgrade/4-8-to-4-9.html . If you have persistent data, refer to the Persistent Data Migration section in the documentation: https://www.hivemq.com/docs/hivemq/4.9/upgrade/4-8-to-4-9.html#persistent-data-migration. If you are using Native SLL TLSv1.1 or TLSv1, refer to the Native SSL section in our documentation: https://www.hivemq.com/docs/hivemq/4.9/upgrade/4-8-to-4-9.html#persistent-data-migration.
We strongly recommend thoroughly testing the upgrade process in the staging environment first and to make sure that the cluster in the new version is working as expected, all clients can communicate to the application successfully, all clients are distributed evenly between the nodes of the cluster, all nodes are evenly loaded. If you are using any custom extension(s) their behavior and functionality in the new version of the cluster also must be thoroughly tested and verified.