autonomy-orchestrator fleet-state rebuild

Replay all events and rebuild the fleet_nodes derived state table

Synopsis

Reads every ingested event in insertion order and replays them through
the fleet-state processor, then atomically replaces the fleet_nodes table.

Use this command to repair the derived state after:
  - manual database edits
  - a crash before a commit was flushed
  - upgrading from a version that did not yet have fleet_state support

The entire operation (event scan + fleet_nodes replacement) executes inside
a single SQLite transaction.  The store uses a single shared connection
(SetMaxOpenConns(1)), so concurrent Ingest calls queue behind the rebuild
transaction rather than racing with it.  The resulting fleet_nodes table is
snapshot-consistent with the events table at transaction-open time.

Usage

autonomy-orchestrator fleet-state rebuild [flags]

Options

      --data-dir string   data directory containing events.db (default: XDG_CACHE_HOME/autonomyops/orchestrator)

See also