http://www.slony.info/bugzilla/show_bug.cgi?id=81

Jan Wieck <janwieck at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|2.0                         |devel

--- Comment #7 from Jan Wieck <janwieck at yahoo.com> 2010-08-24 07:40:10 PDT ---
I agree with Steve that this is covered by 132.

I am moving this to devel since there is still an issue. It is possible that
the OS's PID "almost" wrapped around when the slon worker process dies and that
another process gets its PID reassigned before the cleanup of sl_node_lock
happens. 

The solution to that however is too much change for 2.0, so we will do this in
2.1. The idea for this is to replace sl_node_lock as it is now with a temp
table. Temp tables are visible in pg_class, so an existing sl_node_lock temp
table indicates an existing worker backend connection. If after creating the
temp table there are more than one entry in pg_class, we had a race condition
and need to retry the whole thing.

-- 
Configure bugmail: http://www.slony.info/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
You are the assignee for the bug.