Technologies/Meilisearch/meilisearch.index.documents

MeilisearchMetric

meilisearch.index.documents

Documents in specific index

Dimensions:None

Available on:

Prometheus (1)Native (1)

Interface Metrics (2)

Prometheus

meilisearch_index_documents

Number of documents in a specific index

Dimensions:None

Native

index.number_of_documents

Number of documents in the index

Dimensions:None

Sources

meilisearch_index_documentsdocs.meilisearch.com

index.number_of_documentswww.meilisearch.com

Technical Annotations (45)

Configuration Parameters (4)

searchableAttributesrecommended: minimal set of fields requiring full-text search

Reducing searchable attributes limits FST size and prevents LMDB value size overflow

filterableAttributesrecommended: fields requiring exact match only

Use for fields that don't need full-text search to reduce prefix database bloat

distinct-attributerecommended: original document ID field

prevents duplicate results when documents are split into chunks

client_request_timeoutrecommended: Duration::ZERO (disable) or >5s

Actix timeout that triggers HTTP 408 when workers are unavailable

Error Signatures (11)

MDB_TXN_FULL: Transaction has too many dirty pages - transaction too bigerror code

internalerror code

panicked at milli/src/search/facet/facet_distribution.rs:215:22: called `Option::unwrap()` on a `None` valueexception

500http status

I/O error (os error 5)error code

Input/output error (os error 5). This error generally happens when you have no space left on device or when your database doesn't have read or write right.

log pattern

io_errorerror code

missing_document_iderror code

Document does not have a :primaryKey attributelog pattern

408http status

Expected response code 200, got 408log pattern

CLI Commands (3)

curl -H "Authorization: Bearer ${MEILISEARCH_API_KEY}" -H "Content-Type: application/json" "${MEILISEARCH_HOST}/indexes/${INDEX_NAME}/documents/delete" -d "{\"filter\": \"timestamp < $(date --date=\"00:00 today\" -u \"+%s000\")\"}"

diagnostic

df data.msdiagnostic

curl -X PUT 'MEILISEARCH_URL/indexes/INDEX_NAME/settings/distinct-attribute' -H 'Content-Type: application/json' --data-binary '"story_id"'

remediation

Technical References (27)

tmpfsconcept/tmpfile pathstateful podsconceptLMDBcomponentfacet databasecomponentdocumentDeletioncomponentmilli/src/search/facet/facet_distribution.rsfile path/multi-searchcomponentFSTconceptprefix databasecomponentdata.msfile pathdocumentAdditionOrUpdatecomponentbatch insertconceptHNSWconceptbinary quantizationconceptSSD NVMecomponentseparatorconcepthard separatorconceptproximityconcept#3866componentArroycomponentdistinct attributeconcepttaskUidcomponentprimaryKeyconcept/statscomponentactix-webcomponentactix worker threadscomponent

Related Insights (15)

Docker volume and tmpfs misconfiguration prevents large-scale indexingwarning

▸

Index corruption when bulk deletion fails with MDB_TXN_FULL transaction limitcritical

▸

Multi-search endpoint panics after index corruption with facet distribution errorcritical

▸

I/O error (os error 5) when index exceeds ~800GB due to LMDB value size limitcritical

▸

Large batch document inserts fail at scalewarning

▸

Vector index indexing slows significantly with tens of millions of documentswarning

▸

Separator configuration changes in v1.3.0 increase word proximity calculationsinfo

▸

Indexation time degrades with database sizewarning

▸

High percentage of searches return no resultswarning

▸

Incremental indexing enables sub-minute updates on million-scale vector databasesinfo

▸

Insufficient RAM for dataset size causes memory exhaustioncritical

▸

Large documents reduce search result relevancywarning

▸

Document indexing failure prevents search resultscritical

▸

Sustained high CPU usage with large document countswarning

▸

HTTP 408 errors with large indexes under concurrent loadwarning

▸