Elasticsearch

Cross-Cluster Search Coordination Overhead

warning
latencyUpdated Mar 2, 2026

Cross-cluster search queries can cause significant coordination overhead, network traffic, and timeout issues when remote clusters are slow or have high latency.

How to detect:

elasticsearch.search.query showing elevated latency with high elasticsearch.node.cluster.io and elasticsearch.transport.tx_size when cross-cluster search is active

Recommended action:

Monitor cross-cluster search performance via search profiling and _remote/info API. Optimize: (1) Minimize data transferred - use size and fields parameters aggressively, (2) Use index patterns to limit remote shard count, (3) Implement skip_unavailable:true to avoid blocking on slow clusters, (4) Set appropriate timeouts for remote clusters, (5) Use minimize_roundtrips:true to reduce network hops. For frequently accessed remote data, consider replication or periodic snapshots to local cluster. Monitor network latency between clusters and ensure adequate bandwidth. Review elasticsearch.cluster.in_flight_fetch for concurrent remote operations.