Elasticsearch: Yellow Status
Understand and resolve Elasticsearch Yellow status.
An Elasticsearch cluster in Yellow status means all primary shards are allocated, but some replicas are not. The cluster works but without complete redundancy.
Symptoms
- Cluster health: yellow in _cluster/health
- Unassigned shards reported
- Correct performance but warnings
- No redundancy for certain indexes
Common Causes
- Single node: With a single node, replicas cannot be allocated.
- Insufficient disk space: ES refuses to allocate shards if disk is > 85% full.
- Node down: A cluster node is down and replicas are waiting.
Diagnostic Steps
- Check _cluster/health for details
- List unassigned shards with _cat/shards
- Examine disk space on nodes
- Check node state (_cat/nodes)
Automate with MoniTao
MoniTao can monitor your Elasticsearch:
- Monitoring of _cluster/health endpoint
- JSON content validation for status
- Alerts if status is not green
Best Practices
- Use at least 3 nodes for a production cluster
- Configure index.number_of_replicas according to your nodes
- Monitor disk space (< 80%)
- Set up alerts on cluster health
FAQ
Is Yellow critical?
No, the cluster works. But you don't have complete redundancy.
How to force replica allocation?
Add nodes or reduce number_of_replicas to 0 (not recommended).
What minimum disk space?
ES blocks allocations > 85% and goes read-only > 95%.
Does MoniTao monitor ES natively?
Via HTTP monitoring of _cluster/health with JSON validation.
Useful Links
Ready to Sleep Soundly?
Start free, no credit card required.