Indices Without Replicas
One or more indices are configured with no replica shards. If a primary fails or data is corrupted, there is no promoted copy to keep the index online.
For a complete list of insights, refer to AutoOps insights.
| Field | Value |
|---|---|
| Component | Elasticsearch |
| Severity | High |
| Scope | Index |
| Domains | performance, cluster-health, replication, stability |
You can customize these settings to adjust when AutoOps detects this event and presents the insight. Refer to AutoOps event settings for details.
The default customization settings are:
| Setting | Type | Default |
|---|---|---|
| Exclude indices based on this pattern | List of strings | |
| Data roles (tiers) to exclude from indication | List of strings | |
| Don’t raise event if the index is smaller than (MB) | Integer | 0 |
| How long after index creation do you want to postpone this event trigger? (In min) | Integer | 0 |
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.
Total number of indices without replicas: 12 .
See indices sample list: logs-prod-000045, logs-prod-000046.
AutoOps shows different recommendations depending on how their conditions match your deployment or cluster.
Add data node
Condition: Shown when only one data node available.
Add a data node to increase capacity and reduce pressure on the existing nodes.
Increase replica count
Condition: Shown when index has no replica shards and enough data nodes are available.
Set number_of_replicas to 2 on logs-prod-000045 (currently 1) using the action below.
PUT logs-prod-000045/_settings
{
"index": {
"number_of_replicas": 2
}
}
Requires the manage index privilege. Requires Elasticsearch 8.0.0 or later. This action changes cluster or index configuration.
Impact: Indices without replicas can lead to data loss or search latency. Replicas are important for two reasons: Firstly, they provide resilience: in the event of data corruption or loss of a data node, a replica shard would be promoted to become the primary shard, so that the cluster can continue to run searches, and add or update data in the index. If you have no replicas then this will not happen, meaning that in the event of corruption or loss of the shard, at best you will have to wait for the shard or node to recover, at worst the data might be irrecoverable and you will need to restore the index from a snapshot (if available). The second function of replicas is to provide alternative sources of data to balance the load of search operations across the nodes in the cluster. In the absence of replicas, when a node is stressed, any search operations requiring shards on that node will be required to wait for the node to respond since they have no alternative sources of data.