AI search engine and MCP server for bots and agents. One Rust binary indexes code, docs, logs, JSONL/CSV and PDFs for BM25 full-text, vector and hybrid search, code search without grep, agent memory and RAG. Elasticsearch-compatible API; Corpus Hub serves hundreds of reference memory corpora.
Rust agent-memoryai-agentsbm25claude-codeclicode-searchcursordatabaseelasticsearchfull-text-searchhybrid-searchjsonllocal-firstmcpmcp-servermodel-context-protocolragrustsearch-enginevector-search
test(reindex): the >10k keyset test was measuring the runner's core count
xerj-org2888e834fe57
diff --git a/engine/crates/xerj-api/src/es_compat.rs b/engine/crates/xerj-api/src/es_compat.rsindex 3d1199512..f736f9dd4 100644--- a/engine/crates/xerj-api/src/es_compat.rs+++ b/engine/crates/xerj-api/src/es_compat.rs@@ -30988,6 +30988,17 @@ mod reindex_keyset_tests {n as u64,"sanity: all source docs indexed");+ // `live_doc_count` reads the version map, but reindex pages the SEARCH+ // surface, and turbo ingest with `parallel = true` scatters docs across+ // memtable shards by worker thread. Without this refresh the assertion+ // below silently measures how many cores the machine has: on a 2-core+ // runner every doc lands in one shard and happens to be visible, so CI+ // saw 10050/10050 and went green, while any developer box with real+ // core count searched an unpublished memtable and got 0. The HTTP+ // `_reindex` path is unaffected — verified end-to-end at 10050/10050+ // both with and without an explicit `_refresh` — so this is the test+ // reaching under the API, not a product defect.+ src.refresh().await.expect("refresh src");// Reindex src -> dst with a 1000-doc page size (forces >10 pages,// crossing the 10k window that a from-offset pager cannot).