> This benchmark do not help much in real life, if one care database
> performance.

Did you even read the article? They used a real-world bookstore
application called Nile written in JSP and tested it with 50 to 1,000
concurrent users.
> Database performance will vary a lot with
> It's very easy to make one is better while another is not.

Yes, that's why they fine tuned each one. They also invited each company
to send representatives to help them tune it. Only MySQL and Sybase did
so, IMB offered suggestions over email and apparently Oracle and
Microsoft couldn't be bothered. 

> I suggest to take benchmark by yourself with schema and query that
> is similar to production system. (Don't forget to test with large
> data)

Yes, perhaps you can loan me the money to get Oracle and MSSQL licenses
so I can test them out?

I understand what you're saying, every application is different. Yes, I
agree, but for those that can't afford to do extensive benchmarking like
this, we have to use these results and assume ours will be the same. 

What's this got to do with PHP? Nada... Happy New Year!

---John Holmes...

PHP Database Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to