# Retrieval Federated Search and Source Routing
Status: public
Confidence: medium (0.725) (verified)
Last verified: 2026-06-02
Generation: ai_structured


## TL;DR

Federated search routes one retrieval request across multiple sources or clusters, then merges evidence for the agent.

## Core Explanation

Agents using RAG often need evidence from more than one index, tenant, region, or product surface. Federated search keeps those sources separate while issuing coordinated queries and merging results.

The operational evidence to inspect includes routed sources, per-source query latency, partial failures, score calibration, duplicate handling, and whether authorization filtering happened before or after fanout.

## Source-Mapped Facts

- Elastic documentation describes cross-cluster search as searching across multiple remote clusters. ([source](https://www.elastic.co/docs/explore-analyze/cross-cluster-search))
- OpenSearch documentation describes cross-cluster search as querying data across multiple connected clusters. ([source](https://docs.opensearch.org/latest/search-plugins/cross-cluster-search/))
- Apache Solr documentation describes distributed requests across SolrCloud collections and shards. ([source](https://solr.apache.org/guide/solr/latest/deployment-guide/solrcloud-distributed-requests.html))

## Further Reading

- [Elastic Cross-Cluster Search](https://www.elastic.co/docs/explore-analyze/cross-cluster-search)
- [OpenSearch Cross-Cluster Search](https://docs.opensearch.org/latest/search-plugins/cross-cluster-search/)
- [SolrCloud Distributed Requests](https://solr.apache.org/guide/solr/latest/deployment-guide/solrcloud-distributed-requests.html)