when you attempt to update cluster twice, status might get stuck.
Then delete tasks that you see are conflicting
It could also be that you manually create revisions and aws is trying to replace those with new revisions. If so, deregister the old revisions and let the update will finish alot faster