This topic describes the release notes for Realtime Compute for Apache Flink and provides links to relevant references. The release notes provide the major updates and bug fixes in Realtime Compute for Apache Flink in the version that was released on January 11, 2021.
Major updates on the platform
Intelligent diagnosis is supported. This feature is used to automatically diagnose errors that occur in the deployments that are running.
The metrics of a deployment in a session cluster can be viewed.
JobID can be used to identify a deployment.
A session cluster can be scaled in.
Authorization by using a RAM user is supported on Ververica Platform (VVP).
A deployment can be switched between a session cluster and a per-job cluster.
The maximum size of a JAR package that you can upload on VVP is increased to 200 MB.
NoteIf the JAR package exceeds 200 MB, we recommend that you split the package or upload the package by using Object Storage Service (OSS).
Major bug fixes on the platform
The issue of slow syntax check is fixed.
Major updates in VVR 2.1.3
A JAR package can be uploaded for a deployment in a session cluster in Ververica Runtime (VVR).
The automatic tuning policy Source-Delay-Based supports Kafka.
Connector formats are added, which are the same as the connector formats that are defined in the Apache Flink community.
Metrics including the latency are added. For more information, see Metrics.
Major bug fixes in VVR 2.1.3
The issue that an error occurs because the Change Data Capture (CDC) connector does not support the rename operation on PolarDB tables is fixed.
The issue that NullPointException occurs or restoration based on a savepoint fails due to defects in the Gemini state backend is fixed.