On 4 Jul 2003 at 16:35, Michael Mattox wrote:
> I see this as a major problem. How many people run postgres, decide it's
> too slow and give up without digging into the documentation or coming to
> this group? This seems to be pretty common. Even worst, they tell 10
> others how slow Postgres is and then it gets a bad reputation.
>
> In my opinion the defaults should be set up for a typical database server
> machine.
Well, there are few major reasons defaults are the way they are and the reason
it hurts the way they are
1. Postgresql expects to start on every machine on which it can run. Now some
of the arcane platforms need kernel recompilation to raise SHMMAX and defaults
to 1MB.
2. Postgresql uses shared memory being process based architecture. Mysql uses
process memory being threaded application. It does not need kernel settings to
work and usually works best it can.
3. We expect users/admins to be reading docs. If one does not read docs, it
does not matter what defaults are. Sooner or later, it is going to fall on it's
face.
4. Unlike likes of Oracle, postgresql does not pre-claim resources and starts
hogging the system, replacing OS whereever possible. No it does not work that
way..
One thing always strikes me. Lot of people(Not you Michael!..:-)) would
complain that postgresql is slow and needs tweaking are not bothered by the
fact that oracle needs almost same kind of and same amount of tweaking to get
somewhere. Perception matterrs a lot.
I would have whined for java as well but this is not the forum for that..:-)
On a positive note, me and Josh are finishing a bare bone performance article
that would answer lot of your questions. I am counting on you to provide
valuable feedback. I expect it out tomorrow or on sunday..Josh will confirm
that..
Bye
Shridhar
--
Theorem: a cat has nine tails.Proof: No cat has eight tails. A cat has one tail
more than no cat. Therefore, a cat has nine tails.
