Re: [tlug] MySQL vs Oracle

[Matt Doughty <mdoughty@example.com>]

On Wed, Jun 05, 2002 at 06:55:04PM +0900, Stephen J. Turnbull wrote:
> >>>>> "Matt" == Matt Doughty <mdoughty@example.com> writes:
> 
>     Matt> Is this still the case? I was under the impression MySQL
>     Matt> with InnoDB fixes both the locking issues, and the ACID
>     Matt> compliance.
> 
> I don't know.  That is, yes, it greatly improves the locking issue by
> giving better granularity, I dunno if InnoDB gives ACID compliance (it
> did _not_ the last time I looked, like 12 months ago, but it probably
> does now), BUT I would bet it also fixes that "raw speed problem" that
> MySQL has.  ;-)

I'll save you some time since this made me curious, and the answer is
yes it is ACID compliant:

http://www.innodb.com/features.html

I can't say about the speed. I looked around and I couldn't find any
benchmarks I considered conclusive (just typical single transactions 
times against PG with apologistic overtones)

> 
> I don't need more than PG gives me, and in a brief trawl across the
> web I saw a lot of people _ask_ that question, and nobody answered.  I
> suspect that basically nobody is using InnoDB ;-).  It's worth taking
> a look around, when I get a round tuit or so....
> 

We have postgres on all the servers here, and it has worked beutifully
so I don't have first hand experience with MySQL and don't really feel
the need to explore. :)

--Matt