Re: TODO Done. Superuser backend slot reservations - Mailing list pgsql-hackers

From Tom Lane
Subject Re: TODO Done. Superuser backend slot reservations
Date
Msg-id 22640.1030370448@sss.pgh.pa.us
Whole thread Raw
In response to Re: TODO Done. Superuser backend slot reservations  ("Nigel J. Andrews" <nandrews@investsystems.co.uk>)
Responses Re: TODO Done. Superuser backend slot reservations  (Bruce Momjian <pgman@candle.pha.pa.us>)
List pgsql-hackers
"Nigel J. Andrews" <nandrews@investsystems.co.uk> writes:
> I was taking the line that the last slots in the array are
> reserved. Those are not going to be taken by non su connections.

But that doesn't do the job, does it?  My view of the feature is that
when there are at least MaxBackends - ReservedBackends slots in use (by
either su or non-su connections) then no new non-su jobs should be let
in.  For example, if the system is full (with a mix of su and non-su
jobs) and one non-su job quits, don't we want to hold that slot for a
possible su connection?

Your approach does have the advantage of being very cheap to test
(I think my semantics would require counting the active backends),
but I'm not sure that it really does what we want.

            regards, tom lane

pgsql-hackers by date:

Previous
From: "Marc G. Fournier"
Date:
Subject: Re: [GENERAL] PostgreSQL 7.2.2: Security Release
Next
From: Neil Conway
Date:
Subject: Re: @(#)Mordred Labs advisory 0x0006: Two minor DoS conditions in PostgreSQL