Skip to main content

Search Engine

The Administration → Search Engine page monitors the Elasticsearch cluster that powers TestPlanIt's search, lets you tune the replica count, and rebuilds (reindexes) search indices when needed.

note

Only system administrators can open this page.

How to access​

  1. Open the Admin area from the top navigation.
  2. Select Search Engine under System in the admin menu.

Status​

The status card shows the live connection and health of the cluster:

  • Connection — Connected or Disconnected.
  • Cluster health — a GREEN, YELLOW, or RED badge.
  • Nodes — the number of nodes in the cluster.
  • Indices — a per-index list of the testplanit-* indices with each index's document count, store size, and health.

Use Refresh to re-check the status. If Elasticsearch is not configured or not responding, the card shows a disconnected state and the reindex controls are unavailable.

Configuration​

The Number of Replicas setting controls index redundancy:

  • 0 — for single-node clusters. This is also the fix when health shows YELLOW on a single node (replicas can't be allocated with only one node).
  • 1 or more — for multi-node clusters that need redundancy.

The value is validated to the range 0–10. Saving persists the setting and applies it live to all existing testplanit-* indices.

info

The Configuration card is hidden in multi-tenant deployments, where Elasticsearch settings are shared across tenants and managed at the deployment level.

Reindexing​

Reindexing rebuilds search indices — for example after changing index mappings or to recover from missing documents. Choose a scope and click Start Reindex:

  • All Entities, or one of: Repository Cases, Test Runs, Sessions, Shared Steps, Issues, Milestones, Projects.

Progress, a live log, and a completion summary appear while the job runs.

warning

Reindexing deletes and recreates each target index before repopulating it, so search results may be incomplete while the job runs. A reindex can take several minutes depending on data volume. Run it during a maintenance window when possible.