Anthony shows you how to
fix issues with
red Elasticsearch clusters

anthony_add_storage_elasticsearch

I've run out of storage space in my Amazon Elasticsearch Service (Amazon ES) domain. What does this mean, and how do I fix it?

When you allocate storage to an Elasticsearch node, up to 20% of that space (20 GB) is reserved space. If your domain runs out of storage space, the operation of your domain might be impaired. To increase the available storage space in your Amazon ES domain, try the following:

  • Add more storage space to your domain by increasing the size of its Amazon Elastic Block Store (Amazon EBS) volumes or adding additional nodes to your cluster.
  • Reduce the amount of data stored in your domain by deleting unnecessary files, optimizing old indices, or reducing the domain's replica count.

Add more storage to your domain

If you're using Amazon EBS as the storage solution for your domain, when you run out of storage, you receive a ClusterIndexWritesBlocked error. To add additional storage, increase the size of the EBS volumes used by your domain or cluster.

Note: Each node has a size limit for the EBS volumes that can be attached to it. If you can't attach more storage to your node, add additional nodes or scale up your domain.

If you're not using Amazon EBS as the storage solution for your domain, add additional nodes to your cluster configuration.

Note: The FreeStorageSpace CloudWatch metric measures the number of free megabytes of storage for the nodes in your cluster. Configure CloudWatch alarms to notify you when the available storage for your domain or cluster is low.

CloudWatch metrics report cluster level stats. To check the node level disk allocation, use the cat/allocation command, similar to the following:

curl ES_Endpoint/_cat/allocation?v

Reduce the amount of data stored in your domain

Take a manual snapshot of the domain, and then delete unnecessary documents.

Additionally, you might optimize old indices or reduce the domain's replica count. Note that optimizing old indices might not result in significant amounts of reclaimed storage, and reducing the domain's replica count might decrease the domain's overall fault tolerance.


Did this page help you? Yes | No

Back to the AWS Support Knowledge Center

Need help? Visit the AWS Support Center

Published: 2016-11-01

Updated: 2018-07-26