Indexing on non-hot nodes
Significant indexing activity is running on nodes outside the hot tier. Warm and cold tiers are optimized for retention cost, not ingest throughput.
For a complete list of insights, refer to AutoOps insights.
| Field | Value |
|---|---|
| Component | Elasticsearch |
| Severity | Medium |
| Scope | Cluster |
| Domains | performance, indexing, index-management |
The following is an example of what you might see when this insight is triggered. Real insights use live data and links from your deployment or cluster.
Indexing rate on non-hot data nodes:
- logs-prod-000045: 1,200 docs/s
Indices with the highest indexing rate: 12
AutoOps shows different recommendations depending on how their conditions match your deployment or cluster.
Review data tier architecture
Condition: Always shown for this insight.
Review whether your hot, warm, cold, and frozen tiers match data age, query patterns, and cost goals.
Review templates and ILM
Condition: Always shown for this insight.
Review index templates and ILM policies so new write indices stay on the hot tier until rollover is appropriate. Premature moves to colder tiers usually mean the ILM policy is misconfigured.
Move shard to hot tier
Condition: Shown when index on non-hot tier can be moved to hot nodes that can host the index shards.
Move shard 0 of index logs-prod-000045 from warm node es-data-01 to hot node es-data-02 using the action below.
POST _cluster/reroute
{
"commands": [
{
"move": {
"index": "logs-prod-000045",
"shard": 0,
"from_node": "es-data-01",
"to_node": "es-data-02"
}
}
]
}
Requires the manage cluster privilege. Requires Elasticsearch 8.0.0 or later. This action changes cluster or index configuration.
The hot data tier is designed for active writes and recent search. Warm, cold, and frozen tiers trade ingest performance for lower storage cost and longer retention.
When active indexing runs on non-hot nodes, write latency can rise and tier resources meant for read-heavy workloads are consumed by ingest. This often points to an ILM policy that moves indices before writes finish, or shard allocation rules that place writable shards on the wrong tier.
If the pattern continues, ingest performance can degrade and lifecycle transitions might not match your retention goals.
Elasticsearch data tiers: hot, warm, cold, and frozen storage explained