Skip to content

Commit

Permalink
build: fixes bugs in both jar and docker packaging (#213)
Browse files Browse the repository at this point in the history
* due to use of ancient hadoop which uses sun. packages, we have to use
  a full JRE. We can revisit once off that.
* when we updated from spark 2.x to 3.x we missed a maven shade
  expression which led to classes not found in Elasticsearch (others ok)

Signed-off-by: Adrian Cole <[email protected]>
  • Loading branch information
codefromthecrypt authored Dec 15, 2023
1 parent 08fe63f commit e1125e2
Show file tree
Hide file tree
Showing 2 changed files with 4 additions and 2 deletions.
4 changes: 3 additions & 1 deletion docker/Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -44,7 +44,9 @@ ENV VERSION=$version
ENV MAVEN_PROJECT_BASEDIR=/code
RUN /code/build-bin/maven/maven_build_or_unjar io.zipkin.dependencies zipkin-dependencies ${VERSION}

FROM ghcr.io/openzipkin/java:11.0.21_p9-jre as zipkin-dependencies
# Try -jre again once Spark dependencies are up to date. Spark 3.0 uses old Hadoop in the
# Elasticsearch driver which needs com.sun.security.auth.module.UnixLoginModule.
FROM ghcr.io/openzipkin/java:11.0.21_p9 as zipkin-dependencies
LABEL org.opencontainers.image.description="Zipkin Dependencies Aggregator on OpenJDK and Alpine Linux"
LABEL org.opencontainers.image.source=https://github.com/openzipkin/zipkin-dependencies

Expand Down
2 changes: 1 addition & 1 deletion main/pom.xml
Original file line number Diff line number Diff line change
Expand Up @@ -124,7 +124,7 @@
</filter>
<filter>
<!-- org.elasticsearch.spark.sql.SparkSQLCompatibilityLevel -->
<artifact>org.elasticsearch:elasticsearch-spark-20_${scala.binary.version}
<artifact>org.elasticsearch:elasticsearch-spark-30_${scala.binary.version}
</artifact>
<includes>
<include>**</include>
Expand Down

0 comments on commit e1125e2

Please sign in to comment.