aws cluster_error ai_generated true

OpenSearch cluster status is RED: one or more primary shards are not allocated

ID: aws/aws-opensearch-cluster-red

Also available as: JSON · Markdown
75%Fix Rate
80%Confidence
65Evidence
2023-01-01First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
2 active — — —

Root Cause

OpenSearch cluster is in RED status meaning primary shards are unassigned, usually due to storage pressure or node failures.

generic

Workarounds

  1. 82% success Increase storage volume size and check for disk pressure 
    Check FreeStorageSpace metric in CloudWatch; increase EBS volume size via domain configuration

    Sources: https://docs.aws.amazon.com/opensearch-service/latest/developerguide/handling-errors.html

  2. 80% success Delete old indices or use Index State Management (ISM) policies for lifecycle management 
    Delete indices older than retention period; create ISM policy to auto-delete or move to cold storage

    Sources: https://docs.aws.amazon.com/opensearch-service/latest/developerguide/ism.html

Dead Ends

Common approaches that don't work:

  1. Force delete unassigned shards to turn cluster green 85% fail

    Deleting primary shards causes permanent data loss for those indices

  2. Upgrade instance types to fix red cluster status 65% fail

    If the issue is disk pressure, upgrading instance type without adding storage won't help

Error Chain

Frequently confused with: