> [!info] ParallelFinish > <table><tr><td rowspan="4" width="25%"><img src="assets/wl_ipc_parallelfinish.png"></td><td><b>Wait Category</b></td><td>IPC</td></tr><tr><td><b>Mechanism</b></td><td><code>WaitForParallelWorkersToFinish</code></td></tr><tr><td><b>Operations</b></td><td><a href="Operations/Table/SeqScan">SeqScan</a>, <a href="Operations/ResultSet/HashJoin">HashJoin</a></td></tr><tr><td><b>Resources</b></td><td><a href="Resources/CPU">CPU</a>, <a href="Resources/RAM">Memory</a></td></tr><tr><td colspan="3">This wait event represents the <b>Cleanup Overhead</b> of parallel execution. After all rows have been gathered, the leader backend cannot return the final result set until every worker has safely detached from shared memory and successfully terminated. (Analogized by a supervisor capybara sitting at a desk and checking off a clipboard list as small worker frogs in safety vests carry their toolboxes out the door).</td></tr></table>