ammachado commented on PR #23927: URL: https://github.com/apache/camel/pull/23927#issuecomment-4681902403
> The `mirror.gcr.io/bitnamilegacy/tensorflow-serving:2.19.1` image uses a "legacy" namespace. Worth checking that it has ongoing maintenance for long-term reliability. Thanks for flagging this. The `bitnamilegacy` namespace is indeed deprecated: Bitnami restructured their container catalog in August 2025 and moved all versioned/pinned tags to `bitnamilegacy` with no further updates. The concern is valid. However, there is currently no well-maintained `aarch64` image for TensorFlow Serving from any trusted registry: - The official `tensorflow/serving` image is `amd64`-only (long-standing open issue, no `arm64` support planned). - `emacski/tensorflow-serving`, the main community `aarch64` build, is last published at `2.6.0` and is not actively maintained. - AWS Deep Learning Containers publish a `2.19` `aarch64` image, but it is SageMaker-specific and not suitable here. `bitnamilegacy/tensorflow-serving:2.19.1` is frozen but functional. It is the only option that matches our required version on `aarch64`. We can revisit this when either the official TF Serving project adds `aarch64` support or a community project catches up to 2.19+. The remaining alternative is to apply the `skipITs.aarch64` fix: add `<skipITs.aarch64>true</skipITs.aarch64>` to `camel-tensorflow-serving/pom.xml` and remove the `bitnamilegacy` line from `container.properties`. Please advice. _Claude Code on behalf of Adriano Machado_ -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
