Scale Down Cluster Storage with No Downtime

[Dom]

Is it possible to scale down the storage of my cluster from the console without impacting my workload? I initially set it to 600 GB but am only using 40 GB. My cluster is running on AWS.

[Michael_Lynn]

TL;DR: You can’t scale down storage directly in Atlas (on AWS), but you can migrate to a smaller cluster if needed.

Can you confirm that this is a MongoDB Atlas cluster (as opposed to a self managed cluster running in AWS)? Assuming it’s a MongoDB Atlas cluster, unfortunately, you are unable to reduce the amount of storage in use… see this doc. You can deploy a separate cluster with a reduced amount of disk and use MongoDB Live Migration to move the data over.

Steps for Live Migration:

1. Create a new, smaller Atlas cluster (e.g., with something like 100GB instead of 600GB).
2. Set up Live Migration from the old cluster to the new one.
3. Atlas will sync data in real-time until everything is caught up.
4. Switch your app’s connection string to point to the new cluster.
5. Once everything looks good, decommission the old cluster.

To learn more about MongoDB Live Migration - check out this learning byte.

[Dom]

Thank you for answering, but it is definitely possible. Messages like this misleaded me. It is not necessary to compact data or perform migrations, just change the storage size from the Atlas UI.

As confirmed by the technical support: “Changes are applied in a rolling fashion, starting with the secondaries. Making changes to a cluster often requires migrating to new servers and storage volumes. The time required for an initial sync (https://www.mongodb.com/docs/master/reference/glossary/#std-term-initial-sync) and resynchronizing data across storage volumes increases linearly with the amount of data in the cluster”.

The guide you linked says: “Atlas can reduce storage capacity in place. Atlas provisions new volumes then syncs data from the old to the new volumes. This works around the AWS limitation.”

I hope this can be helpful for other people.