Ververica Platform 2.0.2
- This release adds support for Flink 1.8.3, improves the installation process on OpenShift and fixes various bugs and
Ververica Platform 2.0.2 adds our distribution of Apache Flink® 1.8.3 as well as an additional off-cycle bugfix release for Apache Flink® 1.9.1. The following new Docker images are available:
These images come with Apache Flink® bugfixes, improvements to Ververica Platform’s Kubernetes-based
HighAvailabilityServices and numerous security vulnerability fixes.
Please check Ververica Platform Docker Images for all available Apache Flink® images and additional tags.
- Due to a regression in Ververica Platform 2.0, Deployments with restore strategy
LATEST_STATE did behave exactly as
with restore strategy
LATEST_SAVEPOINT. This is now fixed for Deployments using
Flink Master Failover with Kubernetes. We recommend users relying on
LATEST_STATE to switch from Zookeeper to Kubernetes for Flink Master Failover (see also
Requirements for LATEST_STATE Restore Strategy).
- In the course of this bugfix,
LATEST_STATE restore strategy was improved to become more explicit and transparent
for users: as before this restore strategy relies on Apache Flink® retained checkpoints. These are now listed in the
“Snapshots”-tab (formerly “Savepoints”) of the web user interface, just like regular savepoints. By this, users can
now manually reset their Deployments to retained checkpoints, too, and - more importantly - users can always
see if a Deployment will recover from a retained checkpoint when using restore strategy
LATEST_STATE. Please see
Requirements for LATEST_STATE Restore Strategy for details.
- Quotation marks in
Deployment.spec.template.spec.artifact.mainArgs are handled more robustly now. See
Deployment Templates for details.
Please download the latest Helm charts for this version from the customer portal and run
$ helm upgrade [RELEASE] ververica-platform-2.0.2.tgz --values custom-values.yaml