This topic describes the major updates released for Realtime Compute for Apache Flink on March 7, 2024.March 7, 2024
This version rolls out incrementally using a canary release strategy. New features are available only after the upgrade is complete for your account. To check the upgrade schedule, see the announcement on the right side of the Realtime Compute for Apache Flink console. To request an early upgrade, submit a ticket.
Features
Cross-zone high availability
A single-zone outage no longer brings down your deployments. Cross-zone high availability is now supported, letting you create namespaces that use cross-zone compute units (CUs). When a fault occurs in one zone, affected deployments automatically resume in another zone within the same region—without service interruption.
To enable this feature, activate it for your workspace, then create a namespace with cross-zone CUs.
Related topics:
Operator profiling (public preview)
Tracking down incorrect data in a running SQL deployment used to mean adding logging or restarting jobs. Operator profiling is now available for public preview, letting you query intermediate results of a running SQL deployment directly. This makes it significantly faster to verify data correctness without disrupting the running job.
Related topics:
CloudMonitor integration
CloudMonitor is now available as a monitoring service option when you purchase a workspace. It provides basic monitoring and alerting at no charge—you are billed only for notifications sent through SMS and phone calls.
Related topics:
Optimization of the Workflows feature (public preview)
The following improvements are now available for the Workflows feature in public preview:
Failure alerts: Alerts are sent automatically when a task in a workflow fails.
Task-level controls on the task details page: Three operations are now available—Re-Run (rerun a task), Set as Success (mark a failed task as successful), and Force Stop (terminate a task).
Related topics:
Bulk start or cancel deployments
Managing many deployments one by one is time-consuming. On the Deployments page in the development console, you can now start or cancel multiple deployments at a time.