tnks for the hint ;) I'll try something similar here.
Andrea
Josh Berkus wrote:
Andrea,
i'm sorry for the curiosity.... but could you share, if it's possible, this workaround? ;) (if it's not the one you describe at the beginning thread e.g. don't use LIMIT 1)
Well, we actually roped in the pg_locks view to do a "SELECT the first row not already locked for update". Then added some code on the client end for error handling, like race conditions and no rows being returned, both of which happen in production.
---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
