Best Practices with Docker Healthcheck

Docker Engine General
1

In the documentation there is no indication as to the impact of setting the interval down to a lower number ( for example 1 second). Are there any best practices on what a developer building an image should do for the health check interval ?

Thanks
Chase

2

depends on your use case.

  • if your health check is expensive, then you shouldn’t do it too often.
  • If your health check is cheap, you can do it a little more often.
  • also if you need to rely on the health of your service much, then you also could invest more time in checking that it is really healthy.
  • if your service is flaky and fails from time to time, then you probably shouldn’t check too often and also given it several retries.

Health checking everytime exec’s a new process. So you shouldn’t do it too often. I think an interval of less than 10 seconds is too small.