Cluster operations workflows

Indexes and search

1. Review health, document counts, storage, and aliases in the index list.
2. Open details for settings, statistics, and mappings.
3. Use pagination, sorting, visible columns, and aggregations in search.
4. Confirm the index, query, and estimated affected documents before writes.

Bulk operations include read-only toggles, open/close, force merge, clone, reindex, and delete-by-query emptying. Validate high-risk operations against test clusters first.

Shard diagnostics

• Use the shard matrix to find unassigned or relocating shards.
• Run allocation explain for the reason.
• Perform reroute dry-run before reviewing and applying the final request.
• Follow _cat/recovery progress and raw _recovery details.

Snapshots and restore

1. Create and verify a repository.
2. Select indexes, feature states, metadata, and partial-snapshot behavior.
3. Preview matched indexes and rename targets before restore.
4. Resolve target conflicts and review the final request JSON.
5. Run the restore and monitor progress.

Full-index export

Choose a local download or tested S3 profile. Jobs enforce concurrency, queue, retry, and cancellation boundaries. After failure, inspect target capacity, network, credentials, and incomplete multipart uploads before retrying.