Why is my Docker container exiting with code 137 and how can I fix this memory issue?

All answers to this question.

Exit code 137 almost always indicates a SIGKILL caused by an Out-Of-Memory (OOM) event. You should first run docker inspect [container_id] and look for the OOMKilled flag. If it's true, Docker or the host kernel killed the process because it exceeded its memory quota. To fix this, try increasing the memory limit in your docker-compose file or using the --memory flag in your run command. Additionally, check your application for memory leaks, especially if you're processing large datasets in-memory without clearing the buffer.

   Answered 2025-03-16 by Jennifer Taylor

Are you using any specific monitoring tools like Prometheus to track the memory usage spikes before the container exits?

   Answered 2025-03-17 by Michael Harris

• That’s a valid point, Michael. I haven't set up Prometheus yet, but I started using docker stats to watch the memory usage in real-time. I noticed that the memory consumption climbs steadily until it hits the 2GB limit I set, then immediately triggers the 137 error. This confirms it's a resource limit issue rather than a random crash. I'll try doubling the limit to see if it stabilizes.

     Commented 2025-03-18 by Robert Miller

Sometimes code 137 happens if you manually stop a container and the application doesn't handle the SIGTERM within 10 seconds, forcing a SIGKILL.

   Answered 2025-03-19 by Sarah Williams

• Sarah is right. If your app takes too long to shut down, Docker loses patience and kills it. Increasing the stop timeout can often prevent these "false" OOM errors.

     Commented 2025-03-20 by Jennifer Taylor