On 12-09-28 02:40 PM, Brian Fehrle wrote:

> Is there anything I can provide that could help us dig into this more?
> Are we counting this as a bug that may have been fixed in 2.2(haven't
> looked at what you mentioned yet Steve)?

My best guess is that this like the situation that Jan described where 
one thread was processing/adding a SYNC from the dropped node at the 
same time a different thread in the same slon was processing the DROP 
NODE.

My understanding is that the automatic WAIT FOR in 2.1 only waits for 
non-sync events to propogate everywhere before starting with the DROP 
NODE.  I *think* this has been fixed in 2.2 but 2.2 does need more testing.



>
> Thanks,
> - Brian F
>
>>
>>>
>>> Jan
>>>
>>
>