Avoid logging error from updating status of a deleted workspace

When the last finalizer on a DevWorkspace is cleared, the cluster may garbage-collect that workspace immediately, which can result in the object being deleted before we reach the point of attempting to update the workspace's status. This results in logging a harmless but confusing error, so we don't try to update the status for terminating workspaces with no finalizers. Signed-off-by: Angel Misevski <[email protected]>
amisevsk · Jul 4, 2022 · e3805fb · e3805fb
1 parent 8865f7f
commit e3805fb
Showing 1 changed file with 7 additions and 0 deletions.
diff --git a/controllers/workspace/finalize.go b/controllers/workspace/finalize.go
@@ -53,6 +53,13 @@ func (r *DevWorkspaceReconciler) finalize(ctx context.Context, log logr.Logger,
 	finalizeStatus := &currentStatus{phase: devworkspacePhaseTerminating}
 	finalizeStatus.setConditionTrue(conditions.Started, "Cleaning up resources for deletion")
 	defer func() (reconcile.Result, error) {
+		if len(workspace.Finalizers) == 0 {
+			// If there are no finalizers on the workspace, the workspace may be garbage collected before we get to update
+			// its status. This avoids potentially logging a confusing error due to trying to set the status on a deleted
+			// workspace. This check has to be in the deferred function since updateWorkspaceStatus will be called after the
+			// client.Update() call that removes the last finalizer.
+			return finalizeResult, finalizeErr
+		}
 		return r.updateWorkspaceStatus(workspace, log, finalizeStatus, finalizeResult, finalizeErr)
 	}()