CODEWITHSUNDEEP
dockerplayframeworksbtproduction-environment
mrdoubtful
mrdoubtful

Reputation: 529

Unexpected JVM exit after first docker process is completed

I have an application running two services in a docker container. I'm running into a situation where the JVM unexpectedly exits after completing the first process (which is to build a jar from a play project) with the following message:

[info] Your package is ready in /app/.../project-0.1.0.zip
[success]
[Thread-2] [CoordinatedShutdown(akka://sbt-web)] Starting coordinated shutdown from JVM shutdown hook
Archive:  /app/.../target/universal/project-0.1.0.zip

The second process (which is to extract the jar and run the application) begins with Archive: /app/.../target/universal/project-0.1.0.zip but fails to extract the jar. The docker container does not exit.

Surprisingly, this issue is not seen in multiple runs after rebuilding and running the containers. Also, I had recently upgraded my sbt version from 0.13.11 to 1.1.1. It was since then that I started facing this issue of CoordinatedShutdown. I am quite certain thus that this is not a docker issue.

I had also tried adding akka.coordinated-shutdown.exit-jvm = off and akka.coordinated-shutdown.run-by-jvm-shutdown-hook = off to my application.conf but CoordinatedShutdown still occurs. Any solutions?

Upvotes: 0

Views: 1140

Answers (1)

mrdoubtful
mrdoubtful

Reputation: 529

I figured out the issue was that the play application failed to start in its container because the RUNNING_PID file had not been deleted. Finally, managed to solve this by preventing Play from creating a PID with -Dpidfile.path=/dev/null.

Related issues: Restarting Play application Docker container results in 'This application is already running' - RUNNING_PID is not deleted

References: https://www.playframework.com/documentation/2.6.x/Deploying#play-pid-configuration

Upvotes: 1

Related Questions