Skip to main content
Version: 8.9 (unreleased)

Schema and data migration

The orchestration cluster persists runtime and task data in secondary storage.

This page describes schema and migration behavior for document-store secondary storage (Elasticsearch/OpenSearch), where indices and templates are created automatically on first startup. For RDBMS secondary storage guidance, see RDBMS configuration.

Schema​

For document-store backends, cluster data is stored in indices governed by a schema version. Index names follow this pattern:

{cluster-index-prefix}-{legacy-prefix}-{datatype}-{schemaversion}_[{date}]
  • cluster-index-prefix – Prefix for index names (default: ``).
  • legacy-prefix – Legacy component prefix (e.g. operate, tasklist, camunda).
  • datatype – Identifies the type of data stored (e.g., user, variable, task).
  • schemaversion – Version of the index schema.
  • date – Optional date for archived data.

The version in the index name is specific to the schema and may differ from the cluster software version. See data retention.

For more information about the specific indices used by the orchestration cluster, refer to the Index Diagrams.

From 8.8 onwards, no schema migrations are required when upgrading the orchestration cluster.

See also: Version compatibility checks for the rules governing supported upgrade paths and how schema version metadata is validated.