This is the multi-page printable view of this section. Click here to print.
New and changed in Vertica 24.4
- 1: Health Watchdog
- 2: Upgrading Vertica on Kubernetes
- 3: Sandboxing on Kubernetes
- 4: Save restore points from Kubernetes
1 - Health Watchdog
Health Watchdog
A Health Watchdog has been implemented that automatically detects and blocks queries during bad health conditions on ther server. For more details on Heal Watchdog, see Health Watchdog.
2 - Upgrading Vertica on Kubernetes
Online upgrade
The current online upgrade policy has been renamed to ReadOnlyOnline, and we now have a new approach for online upgrades. This updated mechanism ensures the database remains active and capable of continuous writes through replication. Additionally, we implemented session transfer to further improve the upgrade experience.
For more information about the upgrade policies, see Upgrading Vertica on Kubernetes.
3 - Sandboxing on Kubernetes
Subcluster sandboxing on Kubernetes
Sandboxing in Kubernetes lets you quickly set up separate testing environments without the need to create a new database or reload data. This makes it easier to test Vertica features in new versions. You can stay connected to the main cluster while the upgrade is done in the sandbox. Once the upgrade is finished, you can easily bring the sandbox back into the main cluster.
For more information about sandboxing, see Sandboxing on K8s.
4 - Save restore points from Kubernetes
Save restore points using VerticaDB Operator
Previously, you could save a restore point using vsql in Vertica, but now there's an alternate method using the VerticaDB operator in Kubernetes, which is more aligned with Kubernetes environments. For more details on creating an archive and saving a restore point, see VerticaRestorePointsQuery custom resource definition.