0

We have an application running in a Docker container, within Kubernetes, but because of certain limitations we can't change the application to log to stdout & stderr. This means when the app fails and triggers a container restart we lose any logs that could help us diagnose the issue. The application writes to a directory path within the container.

Two possibilities that we are wanting explore:

  • Having the directory being written to be a mounted persistent volume, that we can look at even when the container gets restarted
  • Have the latest log file created (rolling logs) be tailed and redirected to stdout

Right now I am wanting to explore the first of the two options, since it should allow us to see the failure state even once the container gets restarted. I will explore the second option in parallel.

Any suggestions how to approach the first option? We are using Digital Ocean, and the latest version of Kubernetes, if that changes anything.

Improve this question

1 Answer 1

Reset to default
0

The application writes to a directory path within the container.

So, mount a volume there. Problem solved. This is no different than any other volume – if you were running docker directly, you'd use -v, and depending on how you administer your specific kubernetes setup, you configure the volumes in a config file, or via some other API.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .