Uh oh!
There was an error while loading. Please reload this page.
[SPARK-47635][K8S] Use Java 21 instead of 21-jre image in K8s Dockerfile - #45761
Closed
dongjoon-hyun wants to merge 1 commit into
Closed
[SPARK-47635][K8S] Use Java 21 instead of 21-jre image in K8s Dockerfile#45761dongjoon-hyun wants to merge 1 commit into
21 instead of 21-jre image in K8s Dockerfile#45761dongjoon-hyun wants to merge 1 commit into
Conversation
dongjoon-hyun
commented
Mar 28, 2024
MemberAuthor
Hi, @viirya . Could you review this bug fix for Apache Spark 3.5.0 and 3.5.1? |
21 instead of 21-jre image in K8s Dockerfileviirya
approved these changes
Mar 28, 2024
dongjoon-hyun
commented
Mar 28, 2024
MemberAuthor
Thank you, @viirya ! |
dongjoon-hyun
commented
Mar 28, 2024
MemberAuthor
Merged to master for Apache Spark 4.0.0. |
This was referenced Aug 1, 2024
HyukjinKwon pushed a commit
that referenced
this pull request
Aug 6, 2024
### What changes were proposed in this pull request? This PR aims to fix `docker-image-tool.sh` to be up-to-date. ### Why are the changes needed? Apache Spark 4 dropped Java 11 support. So, we should fix the following. - #43005 ``` - - Build and push Java11-based image with tag "v3.4.0" to docker.io/myrepo + - Build and push Java17-based image with tag "v4.0.0" to docker.io/myrepo ``` Apache Spark 4 requires JDK instead of JRE. So, we should fix the following. - #45761 ``` - $0 -r docker.io/myrepo -t v3.4.0 -b java_image_tag=11-jre build + $0 -r docker.io/myrepo -t v4.0.0 -b java_image_tag=17 build ``` Lastly, `3.4.0` is too old because it's released on April 13, 2023. We had better use v4.0.0. ``` - $0 -r docker.io/myrepo -t v3.4.0 -p kubernetes/dockerfiles/spark/bindings/python/Dockerfile build + $0 -r docker.io/myrepo -t v4.0.0 -p kubernetes/dockerfiles/spark/bindings/python/Dockerfile build ``` ### Does this PR introduce _any_ user-facing change? No functional change because this is a usage message. ### How was this patch tested? Manual review. ### Was this patch authored or co-authored using generative AI tooling? No. Closes#47618 from dongjoon-hyun/SPARK-49117. Authored-by: Dongjoon Hyun <dhyun@apple.com> Signed-off-by: Hyukjin Kwon <gurwls223@apache.org>
dongjoon-hyun added a commit
to apache/spark-docker
that referenced
this pull request
Sep 18, 2024
### What changes were proposed in this pull request? This PR aims to use JDK for Spark 3.5+ Docker image. Apache Spark Dockerfile are updated already. - apache/spark#45762 - apache/spark#45761 ### Why are the changes needed? Since Apache Spark 3.5.0, SPARK-44153 starts to use `jmap` like the following. - apache/spark#41709https://github.com/apache/spark/blob/c832e2ac1d04668c77493577662c639785808657/core/src/main/scala/org/apache/spark/util/Utils.scala#L2030 ### Does this PR introduce _any_ user-facing change? Yes, the user can use `Heap Histogram` feature. ### How was this patch tested? Pass the CIs. Closes#66 from dongjoon-hyun/SPARK-49701. Authored-by: Dongjoon Hyun <dongjoon@apache.org> Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
dongjoon-hyun added a commit
that referenced
this pull request
Mar 21, 2026
…kerfile ### What changes were proposed in this pull request? This PR changes the default Java base image tag from `21` (full JDK) to `21-jre` (JRE only) in the Spark Kubernetes Dockerfile for Apache Spark 4.2.0. ### Why are the changes needed? Since Apache Spark 3.5, SPARK-44153 starts to use `jmap` for UI. - #41709 Since Apache Spark 4.0, we used `21` by default. - #45761 Since Apache Spark 4.2, we can use `21-jre` back by default because we removed `jmap` dependency. - #54919 The JRE image is significantly smaller than the full JDK image since it excludes development tools (compiler, debugger, etc.) that are not needed at runtime. This reduces the Spark container image size for Kubernetes deployments. ``` $ docker images IMAGE ID DISK USAGE CONTENT SIZE EXTRA azul/zulu-openjdk:21 4cfeb0cae1b1 651MB 203MB azul/zulu-openjdk:21-jre 27894d1c0e9d 461MB 114MB ``` ### Does this PR introduce _any_ user-facing change? No behavior change. In addition, users who need the full JDK can still override via `--build-arg java_image_tag=21`. ### How was this patch tested? Pass the CIs and manual review of the Dockerfile change. ### Was this patch authored or co-authored using generative AI tooling? Generated-by: Claude Code (claude-opus-4-6) Closes#54940 from dongjoon-hyun/SPARK-56128. Authored-by: Dongjoon Hyun <dongjoon@apache.org> Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
terana pushed a commit
to terana/spark
that referenced
this pull request
Mar 23, 2026
…kerfile ### What changes were proposed in this pull request? This PR changes the default Java base image tag from `21` (full JDK) to `21-jre` (JRE only) in the Spark Kubernetes Dockerfile for Apache Spark 4.2.0. ### Why are the changes needed? Since Apache Spark 3.5, SPARK-44153 starts to use `jmap` for UI. - apache#41709 Since Apache Spark 4.0, we used `21` by default. - apache#45761 Since Apache Spark 4.2, we can use `21-jre` back by default because we removed `jmap` dependency. - apache#54919 The JRE image is significantly smaller than the full JDK image since it excludes development tools (compiler, debugger, etc.) that are not needed at runtime. This reduces the Spark container image size for Kubernetes deployments. ``` $ docker images IMAGE ID DISK USAGE CONTENT SIZE EXTRA azul/zulu-openjdk:21 4cfeb0cae1b1 651MB 203MB azul/zulu-openjdk:21-jre 27894d1c0e9d 461MB 114MB ``` ### Does this PR introduce _any_ user-facing change? No behavior change. In addition, users who need the full JDK can still override via `--build-arg java_image_tag=21`. ### How was this patch tested? Pass the CIs and manual review of the Dockerfile change. ### Was this patch authored or co-authored using generative AI tooling? Generated-by: Claude Code (claude-opus-4-6) Closesapache#54940 from dongjoon-hyun/SPARK-56128. Authored-by: Dongjoon Hyun <dongjoon@apache.org> Signed-off-by: Dongjoon Hyun <dongjoon@apache.org>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
This PR aims to use Java 21 instead of 21-jre in K8s Dockerfile .
Why are the changes needed?
Since Apache Spark 3.5.0, SPARK-44153 starts to use
jmaplike the following.spark/core/src/main/scala/org/apache/spark/util/Utils.scala
Line 2030 in c832e2a
Does this PR introduce any user-facing change?
No.
How was this patch tested?
Pass the CIs and manual review.
Was this patch authored or co-authored using generative AI tooling?
No.