| From: | Mike Nolan <nolan(at)gw(dot)tssi(dot)com> |
|---|---|
| To: | jd(at)commandprompt(dot)com (Joshua D(dot) Drake) |
| Cc: | scrappy(at)postgresql(dot)org (Marc G(dot) Fournier), alex(at)meerkatsoft(dot)com (Alex), postgresql(at)finner(dot)de (Frank Finner), pgsql-general(at)postgresql(dot)org |
| Subject: | Re: PG vs MySQL |
| Date: | 2004-03-29 18:35:23 |
| Message-ID: | 200403291835.i2TIZOqG014932@gw.tssi.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
> > I also wonder how well the pg_hba.conf method will scale. What happens
> > if there are hundreds of client databases or thousands of entries in
> > pg_hba.conf?
>
> Although I personally would like to see a pg_hba table instead of the
> file, I would have to seariously question your implementation if you had
> hundreds of databases on a single machine.
I know of an ISP who has a large number of customers (in excess of 400)
running similar small (probably under 100MB each) MySQL databases. Since
I know each customer has access only to his own data, I assume it is
implemented using a different database for each customer. Whether or not
it is on one or several machines is a detail I'm not sure of.
Without knowing much about how pg uses the pg_hba.conf file, I don't
know what problems porting that ISP to pg might raise, I only cite it
as an example of an extreme case that might not have been anticipated
and thus possibly an inherent limit in the pg_hba.conf method.
I think it is possible to discuss MySQL features in comparison to pg
without getting into an Annie Oakley/Frank Butler-style argument here.
--
Mike Nolan
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Richard Welty | 2004-03-29 18:44:00 | Re: PG vs MySQL |
| Previous Message | Marc G. Fournier | 2004-03-29 18:08:29 | Re: PG vs MySQL |