OpenSearch Cluster Status Yellow

How to fix it

// try these in order, the most common fix is first

1. FIX 1Check why replicas are unassigned

   Copy steps
   1. 01GET _cat/shards?v and look for UNASSIGNED rows with prirep r
   2. 02GET _cluster/allocation/explain
   Did this fix it? Yes No

2. FIX 2Single-node cluster

   Copy steps
   1. 01Add a second node, or
   2. 02Set replicas to 0 for dev: PUT my-index/_settings {"index":{"number_of_replicas":0}}
   Did this fix it? Yes No

3. FIX 3Free disk space

   Copy steps
   1. 01GET _cat/allocation?v to see disk usage per node
   2. 02Delete old indices or add storage
   Did this fix it? Yes No

4. FIX 4Wait if it is transient

   1. 01After a restart or new index the cluster often self-heals as replicas copy
   Did this fix it? Yes No

Symptoms

Health is yellow; data is searchable but has no redundancy for some indices.

Why it happens

• C-01Single-node cluster with number_of_replicas of 1 or more
• C-02Not enough nodes for the replica count
• C-03Low disk space blocking replica placement
• C-04Temporary state after node restart or new index

// this guide also covers

• opensearch cluster yellow
• opensearch yellow status single node
• opensearch replica unassigned

Other OpenSearch errors

• OpenSearch Cluster Status Red
• OpenSearch Unassigned Shards
• OpenSearch High JVM Memory Pressure
• OpenSearch OutOfMemoryError and Circuit Breaker
• OpenSearch Exceeded Maximum Shard Limit
• OpenSearch Index Read-Only: Disk Watermark and ClusterBlockException
• OpenSearch 429 Too Many Requests
• OpenSearch Docker Container Won't Start