Instead of ignoring or only logging errors when loading configuration, the errors shoud be propagated up and handled.
The question of whether to mark vservers as broken or to propagate the error remains unanswered in the internal fork, so I leave it unanswered here as well.
Instead of ignoring or only logging errors when loading configuration, the errors shoud be propagated up and handled.
The question of whether to mark vservers as broken or to propagate the error remains unanswered in the internal fork, so I leave it unanswered here as well.