While connectig to EMR Cluster AWS EMR Automated Failed

Question

My friends are working on Setting Up an EMR Cluster for one of our Applications and since we want the EMR Setup to be resilient, we want to create a StandBy EMR cluster in an AZ different from the Primary EMR Cluster AZ.

However we are not sure how to automate the Failover to the Secondary EMR if the Primary EMR Cluster AZ goes down.

Any help?

Answer 1

Doing that amount of work wont be necessary as EMR by design would issue a new worker or task node if they fail. For the Master node, you have access to all detailed Cloud Watch services as mentioned earlier and one solution would be to set up a Lambda service that provisions a new cluster if your Master Node fails.

Your data persistence is a separate issue but for the most part and by default EMR storage resides on S3 buckets (available to all zones and highest data persistence).