Sure thing, let’s pull up the allocation logs and run a heap snapshot right after that suspicious call. Compare reference counts before and after the update, and we’ll see if any objects are left hanging like a dangling thread. That should point us to the leak.