fbpx

Announcing auto-recovery for Amazon EMR clusters

Elastigroup periodically monitors your cluster to verify that actual capacity = desired capacity

When your cluster’s node count remains lower than desired for a period of time, Spotinst auto-recovery will re-balance your cluster across Spot instance types to ensure your cluster gets back to full capacity.

Amazon EMR lets you easily run big data frameworks, and process vasts amounts of data on EC2 instances. Spotinst Elastigroup EMR provides you with a management layer that lets you run any EMR workload on Spot instances with minimum overhead, and maximum availability.
When running Amazon EMR on Spot instances, workloads periodically get “stuck” in a state where the number of running instances is lower than the desired amount. This can happen for a variety of reasons related to Spot availability and pricing.

Starting today, Elastigroup EMR periodically checks the state of your cluster, and when it is “stuck”, Elastigroup will automatically shuffle your workload across other Spot instance types in order to achieve the desired number of instances in your cluster.

This feature is now released and available to all Spotinst customers in all regions. Visit our console and get started today, or navigate to our docs to learn more.

Aviv Shukron
Product Manager, Elastigroup

Stay current

Sign up for our newsletter, and we'll send you the latest updates on Spotinst, tips, tutorials and more cool stuff!