Ahh, the infamous JDBC benchmark. :-) There's been much commentary on this in the past. There are some things to consider:
1. The native APIs for each DB will be faster. 2. The DB, App Server and Web Server were all on one box. 3. I'm not sure if the MySQL JDBC driver supported the NamedPipeFactory class at the time those benchmarks were taken. If it didn't, then what you are seeing is an upper-bound imposed by I/O restrictions. 4. In MySQL 4.1.1, InnoDB is able to further benefit from the query cache, as it can now use it outside of autocommit mode. 5. This was done on Windows - scalability may be different on different operating systems (Linux 2.6 and FreeBSD 5.2 are likely to be much better). 6. Interestingly, the performance of MySQL dropped by two thirds when the query cache was disabled - this sounds a bit weird as none of the other databases have this sort of mechanism in place yet performed better than 1/3 of the level of MySQL. 7. The JDBC driver has improved in performance since this test, across the board. 8. MySQL 4.1.1 adds vastly improved FULLTEXT capabilities as well as nested queries. These two additions may be of great benefit to some applications (and basing development on MySQL 4.1.x may result in being able to ship around the time 4.1 is declared production ready depending on the development time involved). I'm not sure if all aspects of this benchmark have been discussed. Heikki has said that the performance of MySQL shouldn't have changed between 4.0.0 and 4.0.16 (4.0.16 was the current version when I asked). The fact that the query cache being turned off caused such a large performance drop and that MySQL scaled so closely to Oracle even though the two engines have such different workings points to the limit in this case being the JDBC element of the test. As I've said in other threads, I can't wait for MySQL AB to release their new benchmarks - it will hopefully give us a simple, definitive source for comparison across architectures, operating systems and access methods. Regards, Chris On Mon, 2004-02-16 at 15:08, Daniel Kasak wrote: > Rodrigo Galindez wrote: > > List, > > Is there a site where I can see performance benchmarks on mySQL > > vs. MS SQL Server 2000 ? We plan to migrate a database that's now > > running under SQL Server 2000 to mySQL, because of online hosting > > costs, but our boss is not so sure of doing that, he feels "safe" > > working under SQL Server 2000. > > Sorry is this is a newbie question, this is my first work mySQL. > > Cheers, > http://www.eweek.com/article2/0,4149,293,00.asp > Click on the links for graphs. > > -- > Daniel Kasak > IT Developer > NUS Consulting Group > Level 5, 77 Pacific Highway > North Sydney, NSW, Australia 2060 > T: (+61) 2 9922-7676 / F: (+61) 2 9922 7989 > email: [EMAIL PROTECTED] > website: http://www.nusconsulting.com.au > > ______________________________________________________________________ > -- > MySQL General Mailing List > For list archives: http://lists.mysql.com/mysql > To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED] -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
