Deployment not highly available
Your deployment is running with only one availability zone on its hot tier or warm tier. Without multiple zones and index replicas, a single zone failure can cause downtime or data loss.
For a complete list of insights, refer to AutoOps insights.
| Field | Value |
|---|---|
| Component | Elasticsearch |
| Severity | Info |
| Scope | Cluster |
| Domains | cluster-health |
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.
Your deployment prod-logs has only one availability zone configured for the following tiers: hot.
AutoOps shows different recommendations depending on how their conditions match your deployment or cluster.
Add an availability zone
Condition: Always shown for this insight.
Add a second or third availability zone to each data tier that currently has only one, and configure index replicas for fault tolerance: 1. Open your deployment editor. 2. Under the Hot tier (and Warm tier if present), set availability zones to 2 or 3. 3. Review the configuration and select Save. See Plan for production for availability zone guidance.
High availability requires two complementary settings: multiple availability zones so that your data is distributed across independent infrastructure, and at least one index replica so that each copy of your data resides on a different zone. Configuring extra zones alone is not enough if all shards sit on a single zone, and configuring replicas alone is not enough if those replicas share the same zone. Act on this insight if you cannot accept downtime or data loss between snapshots. Adding availability zones and ensuring replicas are spread across them keeps your data accessible during host-level or zone-level failures and significantly reduces the risk of unrecoverable loss.