OpenSearch
CertifiedTasks that load, search, and manage data in OpenSearch clusters.
Configure OpenSearch hosts with optional basic auth or headers, then run load/bulk/index operations and search or scroll queries against specific indices; routing, refresh policies, HTTP methods, and path prefixes are exposed for fine-grained control.
OpenSearch
Tasks that load, search, and manage data in OpenSearch clusters.
Configure OpenSearch hosts with optional basic auth or headers, then run load/bulk/index operations and search or scroll queries against specific indices; routing, refresh policies, HTTP methods, and path prefixes are exposed for fine-grained control.
tasks
How to use the OpenSearch plugin
Search, index, bulk-load, and query OpenSearch clusters from Kestra flows.
Authentication
Set connection.hosts to a list of OpenSearch node URLs (e.g. ["http://localhost: 9200"]). For basic auth, set connection.basicAuth.username and connection.basicAuth.password. Set connection.trustAllSsl: true for self-signed certificates. Store secrets in secrets and apply connection properties globally with plugin defaults.
Tasks
Search runs a query against one or more indexes — set request as a JSON string or map. Control result handling with fetchType (default FETCH).
Scroll retrieves large result sets using the OpenSearch scroll API — same properties as Search with no size limit per page.
Put indexes a single document — set index and value (a JSON string or map). Optionally set key to control the document ID and opType to choose between index and create semantics.
Get retrieves a single document by index, key, and docVersion.
Request sends a raw HTTP request to OpenSearch — set endpoint (required) and method (default GET). Pass query parameters and a body as needed.
Bulk performs bulk operations from a file in internal storage — set from to a kestra:// URI. Control batch size with chunk (default 1000).
Load bulk-indexes documents from a file — set from, index, and optionally idKey to use a field as the document ID (set removeIdKey: true, the default, to strip it from the indexed document).
Ppl runs an OpenSearch PPL query set in query. Control result handling with fetchType (default FETCH) and response format with format (JDBC by default, also CSV or RAW).
