Elasticsearch
CertifiedTasks that query and write to Elasticsearch clusters for search, indexing, bulk ingestion, and ES|QL.
Configure a shared connection with cluster hosts (and optional basic auth, headers, path prefixes, or trust-all SSL) and supply index names plus search DSL or ES|QL payloads to run. Covers single-document writes and reads, bulk/streaming ingestion, REST requests, and search or scroll operations with controls for routing, refresh policy, and storing or fetching results for downstream steps.
Elasticsearch
Tasks that query and write to Elasticsearch clusters for search, indexing, bulk ingestion, and ES|QL.
Configure a shared connection with cluster hosts (and optional basic auth, headers, path prefixes, or trust-all SSL) and supply index names plus search DSL or ES|QL payloads to run. Covers single-document writes and reads, bulk/streaming ingestion, REST requests, and search or scroll operations with controls for routing, refresh policy, and storing or fetching results for downstream steps.
tasks
How to use the Elasticsearch plugin
Query, index, and bulk-ingest documents across Elasticsearch clusters from Kestra flows.
Authentication
Configure the cluster connection on each task via hosts (a list of host: port entries). For secured clusters, set basicAuth.username and basicAuth.password for HTTP basic auth, or pass an API key via headers as a "Authorization: ApiKey <token>" entry. For self-signed certificates, set trustAllSsl: true — use only in non-production environments. Store credentials in secrets.
Common properties
Set hosts and any auth properties globally with plugin defaults if all tasks in a flow target the same cluster.
Tasks
Search runs a query DSL request against an index and returns matching documents. For large result sets that exceed a single page, use Scroll instead — it pages through all matching documents using the Elasticsearch scroll API and streams them to Kestra internal storage. Esql runs an ES|QL query and is the right choice for SQL-style analytics and aggregations on Elasticsearch data.
For writes, Put indexes or replaces a single document by ID. Bulk performs batched index, update, and delete operations in a single request — use it for any multi-document write to avoid per-document round trips. Load reads records from a Kestra internal storage file and bulk-indexes them, making it the natural follow-on after a download or transform step.
Get retrieves a single document by ID. Request sends a raw HTTP request to any Elasticsearch REST API endpoint — use it for operations not covered by a dedicated task.
