This topic describes the release notes of Realtime Compute for Apache Flink, including updates to the underlying Ververica Runtime (VVR) engine and the end-to-end platform for development, O&M, and management.
Engine updates
VVR 8.0.1 and 8.0.3 occasionally cause data loss in specific scenarios, which affects data accuracy. After careful evaluation, Alibaba Cloud decides to announce the End of Support (EOS) for VVR 8.0.1 and 8.0.3. We recommend that you upgrade to VVR 8.0.5 or later at the earliest opportunity. For information about how to upgrade, see Upgrade the engine version of deployments. We provide necessary support and guidance to help you smoothly transition to a more secure and stable version. Thank you for your understanding and cooperation.
VVR version
| Apache Flink version | General Availability (GA) | End of General Support (EOGS) | End of Full Support (EOFS) | End of Support (EOS) | Release note |
8.0.7 | 1.17.2 | May 29, 2024 | May 29, 2025 | August 29, 2025 | November 29, 2025 | |
8.0.6 | 1.17.2 | April 1, 2024 | April 1, 2025 | July 1, 2025 | October 1, 2025 | |
8.0.5 | 1.17.2 | January 4, 2024 | January 4, 2025 | April 4, 2025 | July 4, 2025 | |
8.0.4 | 1.17.1 | November 27, 2023 | November 27, 2024 | February 27, 2025 | May 27, 2025 | |
8.0.3 | 1.17.1 | October 23, 2023 | - | - | June 14, 2024 | |
8.0.1 | 1.17.1 | August 21, 2023 | - | - | June 14, 2024 | |
6.0.7 | 1.15.4 | June 21, 2023 | To be determined | To be determined | To be determined | |
6.0.6 | 1.15.3 | April 13, 2023 | May 13, 2024 | August 13, 2024 | November 13, 2024 | |
6.0.5 | 1.15.3 | February 20, 2023 | March 20, 2024 | June 20, 2024 | September 20, 2024 | |
6.0.4 | 1.15.2 | December 26, 2022 | January 26, 2024 | April 26, 2024 | July 26, 2024 | |
6.0.3 | 1.15.2 | December 12, 2022 | January 12, 2024 | April 12, 2024 | July 12, 2024 | |
6.0.2 | 1.15.0 | September 19, 2022 | - | - | June 30, 2023 | |
4.0.18 | 1.13.3 | June 21, 2023 | To be determined | To be determined | To be determined | |
4.0.16 | 1.13.3 | November 9, 2022 | - | - | June 30, 2023 | |
4.0.15 | 1.13.3 | September 19, 2022 | - | - | June 30, 2023 | |
4.0.14 | 1.13.3 | August 18, 2022 | - | - | June 30, 2023 | |
4.0.13 | 1.13.3 | May 16, 2022 | June 10, 2023 | - | June 30, 2023 | |
4.0.12 | 1.13.3 | March 4, 2022 | March 10, 2023 | June 10, 2023 | June 30, 2023 | |
4.0.11 | 1.13.3 | December 16, 2021 | December 10, 2022 | March 10, 2023 | June 10, 2023 | |
4.0.10 | 1.13.1 | September 26, 2021 | September 10, 2022 | December 10, 2022 | March 10, 2023 | N/A |
4.0.8 | 1.13.1 | August 4, 2021 | September 10, 2022 | December 10, 2022 | March 10, 2023 |
For information about the naming rule for VVR versions, see Engine version.
For information about the version mapping for Change Data Capture (CDC) connectors between VVR and Apache Flink, see CDC connectors for Apache Flink.
If a VVR version has critical security or stability issues, we will shorten the RLP of the version to three months to avoid severe impact on business caused by this version, and notify the involved users by using the development console of Realtime Compute for Apache Flink or by sending announcements in a timely manner to upgrade the version.
Platform updates
Start date | Major change | Release note |
May 29, 2024 | The compute units (CUs) of an existing namespace can be converted between the single-zone type and the cross-zone type. Operator-level state time-to-live (TTL) is supported in the expert mode of resource configuration. The UI related to data lineage and deployments is optimized. | |
May 6, 2024 | Code search and variable management are supported. The monitoring and alerting feature and usage limits are optimized. | |
March 7, 2024 | The following features are introduced: cross-zone high availability, operator profiling (public preview), CloudMonitor, and startup or cancellation of multiple deployments. | |
December 7, 2023 | The following features are introduced: queue management, Workflows (public preview), and scripts. | |
October 23, 2023 | The automatic tuning capabilities are optimized. Multiple groups of scheduled strategies can be configured in a scheduled tuning plan. Field-level data lineage is supported for SQL deployments. Dynamic parameter updates are supported. You can log on to the Realtime Compute for Apache Flink console by using a RAM role, a member of a resource directory, or a CloudSSO user. | |
June 21, 2023 | Multiple common features on the platform are optimized, such as audit logs, access to Hive clusters that support Kerberos authentication, and monitoring and alerting. The display of specific pages and user experience on these pages are also optimized. | |
April 13, 2023 | Multiple common features on the platform are optimized. The first-level menus in the development console of Realtime Compute for Apache Flink and the process of draft development and deployment O&M are optimized to improve user experience and enhance the alerting capability. | |
December 12, 2022 | Error logs can be queried. You can select information in logs to query the related error in the development console of Realtime Compute for Apache Flink. | |
September 19, 2022 | New APIs are supported on the platform side to integrate the business of users. | |
May 16, 2022 | A built-in Alibaba Cloud documentation is provided. You can directly view the documentation in the development console of Realtime Compute for Apache Flink. Service notices are added to the Realtime Compute for Apache Flink console. This way, you can directly receive various notices including product updates. | |
March 4, 2022 | The ease of use for log viewing and configuration is improved, the Diagnosis tab is optimized, and the template center is added to the development console of Realtime Compute for Apache Flink. |
References
For information about how to view the engine version used by a deployment, see How do I query the engine version of Realtime Compute for Apache Flink that is used by a deployment?
For information about how to upgrade the engine version used by a deployment, see Upgrade the engine version of deployments.
For information about the product lifecycle policy (PLP) of Realtime Compute for Apache Flink and the runtime lifecycle policy (RLP) of the VVR engine, see Lifecycle policies.
For information about the major changes or service notices of Realtime Compute for Apache Flink, see Service notices of Realtime Compute for Apache Flink. You can also log on to the management console of Realtime Compute for Apache Flink to view the information.
Realtime Compute for Apache Flink adopts an enterprise-class state backend named GeminiStateBackend. For information about the performance comparison between GeminiStateBackend and the state backend provided by Apache Flink (RocksDBStateBackend), see GeminiStateBackend.