Re: Performance of the listen command

The world rejoiced as tgl(at)sss(dot)pgh(dot)pa(dot)us (Tom Lane) wrote:
> Flemming Frandsen <ff(at)partyticket(dot)net> writes:
>> I just looked at the pg_listener table:
>> ... and noticed the complete lack of indexen, surely this must be a bug?
>
> No, that was intentional. It's been a long time but I think the
> argument was that the cost of updating the indexes would outweigh
> their usefulness. The listen/notify code is really not designed
> for a great deal of "flap" in the set of active listeners :-(
>
> I'm not particularly concerned about trying to add indexes --- would
> much rather get rid of the table approach entirely. There have been
> prior discussions of this but nothing's been done.

It would be a pretty easy change to add the obvious index; that could
*somewhat* help things for people who use events a lot in the interim.

Consider it stipulated that if lots of NOTIFY requests go through,
that kills a lot of tuples both in table and index...
--
(reverse (concatenate 'string "moc.liamg" "@" "enworbbc"))
http://linuxdatabases.info/info/languages.html
Signs of a Klingon Programmer - 9. "I have challenged the entire
ISO-9000 quality assurance team to a Bat-Leth contest on the
holodeck. They will not concern us again."