"Tom Lane" <tgl@sss.pgh.pa.us> wrote
>
> The real situation is that you must hold at least AccessShareLock on the
> table throughout the entire operation, else you have no defense against
> (say) someone dropping the index or the entire table out from under you.
> And when you add onto this lock in order to lock out writers
> temporarily, you are engaging in a lock upgrade, which can deadlock
> against any sort of exclusive lock request. The fact that you've been
> holding the AccessShareLock for quite a long time means that the window
> for deadlock problems is very wide.
>
Maybe the deadlock problem is solvable, our current deadlock removal
mechanism is like this:
/* who wakes up first who removes himself -- quite unfair :-( */RemoveFromWaitQueue(MyProc);
What if we change to cost-based removal, i.e., remove the one whose cost is
smaller. In this case, an two-days-to-be-done reindex should never get
killed.
Regards,
Qingqing