On Tue, Apr 5, 2016 at 1:10 PM, Simon Riggs <si...@2ndquadrant.com> wrote:
> If a lock is successfully obtained on one table, but not on all tables, it
>> releases that lock and will retry to get them as a group in the future.
>> Since inheritance acts as a group of tables (top + recursive cascade to
>> children), this implementation is necessary even if only a single table is
>> specified in the command.
> I'd prefer to see this as a lock wait mode where it sits in the normal
> lock queue BUT other lock requestors are allowed to queue jump past it.
> That should be just a few lines changed in the lock conflict checker and
> some sleight of hand in the lock queue code.
> That way we avoid the busy-wait loop and multiple DEFERRABLE lock waiters
> queue up normally.
Yeah, that would be better. I can see how to handle a single structure in
that way but I'm not at all certain how to handle multiple tables and
inheritance is multiple tables even with a single command.
X1 inherits from X
There is a long-running task on X1.
Someone requests LOCK TABLE X IN ACCESS EXCLUSIVE MODE WAIT PATIENTLY.
Internally this also grabs X1.
The lock on X might be granted immediately and now blocks all other access
to that table.
There would need be a Locking Group kind of thing so various LockTags are
treated as a single entity to grant them simultaneously. That seems pretty
invasive; at least I don't see anything like that today.