Technologies/Meilisearch/meilisearch.task.duration

MeilisearchMetric

meilisearch.task.duration

Task execution duration

Dimensions:None

Available on:

Prometheus (1)Native (1)

Interface Metrics (2)

Prometheus

meilisearch_task_duration_seconds

Duration of task processing in seconds

Dimensions:None

Native

task.duration

Duration of task execution in ISO 8601 format

Dimensions:None

Sources

meilisearch_task_duration_secondsdocs.meilisearch.com

task.durationwww.meilisearch.com

Technical Annotations (44)

Configuration Parameters (6)

TMPDIRrecommended: /meili_data/tmp or other disk-based path

Points Meilisearch temporary file writes to disk instead of RAM

MEILI_MAX_INDEXING_MEMORYrecommended: 40Gb

Example configuration for 128GB RAM systems during large imports

searchableAttributesrecommended: explicitly define minimal set

Default indexes all fields; explicitly limiting prevents prefix database bloat

filterableAttributesrecommended: use for non-searchable fields

Fields not requiring full-text search should be filterable only to reduce index size

payload_size_limitrecommended: 100MB

Maximum payload size per API call, can be increased if needed

--max-indexing-memoryrecommended: Set to half of desired maximum memory allocation to account for 2x spike

Controls memory budget for indexing but actual usage can reach 2x this value during settings updates

Error Signatures (3)

I/O error (os error 5)error code

io_errorerror code

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

CLI Commands (2)

docker run -d --restart always -p 7700:7700 -e TMPDIR=/meili_data/tmp -v /root/meili_data:/meili_data getmeili/meilisearch:v1.14.0

remediation

meilisearch --log-level=debugdiagnostic

Technical References (33)

ramfsconcept/tmpfile pathTMPDIRcomponentdocumentAdditionOrUpdatecomponentprefix databasecomponentFSTconceptsearchableAttributesconceptLMDBcomponentHNSWconceptbinary quantizationconceptSSD NVMecomponentseparatorconceptproximityconcept#3866componenthard separatorconceptmemory mappingconceptgrenad sortercomponentmillicomponentwrite transactionconceptgrenadcomponentfallocate(2)componentattributesconceptArroycomponenttree nodesconcepttaskUidconcepttask statusconceptsettingsUpdatecomponentTransformcomponentIndexDocumentscomponentSettings::reindexcomponentfilterable attributesconceptIndexStatus::BeingResizedcomponentprepare_for_closingconcept

Related Insights (20)

TMPDIR RAM exhaustion causes I/O errors during large batch indexingcritical

▸

Large batch sizes amplify temporary storage and memory pressurewarning

▸

Default searchableAttributes cause excessive prefix database growthwarning

▸

Vector index indexing slows significantly with tens of millions of documentswarning

▸

Meilisearch v1.38 removes full DB scans during vector indexinginfo

▸

High-density punctuation without spaces causes indexing slowdownwarning

▸

Separator configuration changes in v1.3.0 increase word proximity calculationsinfo

▸

OOM kill when available memory drops below 217 MiB during indexationcritical

▸

Indexation time degrades with database sizewarning

▸

Slow bulk document indexationwarning

▸

Single-threaded bottleneck limits high-core machine performancewarning

▸

Filesystem cache pressure from temporary file proliferationwarning

▸

Index update backlog under high-frequency single-document writeswarning

▸

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

▸

Incremental indexing minimizes write operations to affected nodes onlyinfo

▸

Random vector distributions and IDs create worst-case indexing performanceinfo

▸

Document operations return 202 but changes not visible in search resultswarning

▸

Single document updates cause indexation performance issueswarning

▸

Settings update allocates twice the max-indexing-memory limitwarning

▸

Index unavailable during resize operationinfo

▸