Now when reorg happens (or any other event causing changes of the events in database) we just quit the process. Then on restart we remove stale tree and restart again. Then we restore properly.
We can have better flow of handling cases like that without actual process restart.
Now when reorg happens (or any other event causing changes of the events in database) we just quit the process. Then on restart we remove stale tree and restart again. Then we restore properly.
We can have better flow of handling cases like that without actual process restart.