Re: profiling connection overhead

On Mon, Dec 6, 2010 at 12:57 PM, Josh Berkus <josh(at)agliodbs(dot)com> wrote:
>>> At some point Hackers should look at pg vs MySQL multi tenantry but it
>>> is way tangential today.
>>
>> My understanding is that our schemas work like MySQL databases; and
>> our databases are an even higher level of isolation.  No?
>
> That's correct.  Drizzle is looking at implementing a feature like our
> databases called "catalogs" (per the SQL spec).
>
> Let me stress that not everyone is happy with the MySQL multi-tenantry
> approach.  But it does make multi-tenancy on a scale which you seldom see
> with PG possible, even if it has problems.  It's worth seeing whether we can
> steal any of their optimization ideas without breaking PG.

Please make sure to articulate what you think is wrong with our existing model.

> I was specifically looking at the login model, which works around the issue
> that we have: namely that different login ROLEs can't share a connection
> pool.  In MySQL, they can share the built-in connection "pool" because
> role-switching effectively is a session variable. AFAICT, anyway.

Please explain more precisely what is wrong with SET SESSION
AUTHORIZATION / SET ROLE.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company