In some limited testing with Redis Cluster I've identified the following conditions where a failed master would be unable to failover:

  • The obvious, all the slaves have failed
  • All of a master's slaves on the majority partition are out of sync with their master
  • A master and all of it's slaves are all in a minority partition
  • Used memory is greater than the "maxmemory" config (if keys don't expire)

    We plan on monitoring for these types of conditions, so ops can intervene before a Redis cluster goes unhealthy.

    What other points of failure does a Redis Cluster have?

    Your Answer

    By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

    Browse other questions tagged or ask your own question.