Re: Extensibility of the PostgreSQL wire protocol

On Mon, Feb 22, 2021 at 07:34:51AM -0500, Dave Cramer wrote:
> On Fri, 19 Feb 2021 at 15:39, Álvaro Hernández <aht(at)ongres(dot)com> wrote:
>
> > On 19/2/21 19:30, Jan Wieck wrote:
> > > [...]
> > >
> > > I also am not sure if building a connection pool into a
> > > background worker or postmaster is a good idea to begin with.
> > > One of the important features of a pool is to be able to suspend
> > > traffic and make the server completely idle to for example be
> > > able to restart the postmaster without forcibly disconnecting
> > > all clients. A pool built into a background worker cannot do
> > > that.
>
> Yes, when did it become a good idea to put a connection pooler in
> the backend???

It became a great idea when we noticed just how large and
resource-intensive backends were, especially in light of applications'
broad tendency to assume that they're free. While I agree that that's
not a good assumption, it's one that's so common everywhere in
computing that we really need to face up to the fact that it's not
going away any time soon.

Decoupling the parts that serve requests from the parts that execute
queries also goes a long way toward things we've wanted for quite
awhile like admission control systems and/or seamless zero-downtime
upgrades.

Separately, as the folks at AWS and elsewhere have mentioned, being
able to pretend at some level to be a different RDBMS can only happen
if we respond to its wire protocol.

Best,
David.
--
David Fetter <david(at)fetter(dot)org> http://fetter.org/
Phone: +1 415 235 3778

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate