[EMAIL PROTECTED] wrote:
Hi all, I'm currently experiencing some very strange behaviour at my MySQL server.

Our production server is a Sun-E4500 (8x400 MHz 4 GB Ram running Solaris 8 and Solstice Disksuite...) and a single A5000 "Storedge" with 127 GB disk space (12x9.1GB @ 10K RPM) array connected over a single FC-AL adapter.

The MySQL server we use is 4.0.17 and as all queries such as select and insert are really fast, all "update" queries are extremely slow.

In numbers it is 17 seconds for a single update query on a 500K record table which looks like this :

AGID    | int(10) unsigned  |      | PRI | NULL                | auto_increment
period  | datetime          |      | MUL | 0000-00-00 00":00:00 |
key_     | varchar(100)      |      | MUL |                     |
type_   | tinyint(3)        |      | MUL | 0                   |
name    | varchar(100)      |      |     |                     |
value   | int(100) unsigned |      |     | 0                   |
lastrun | datetime          |      |     | 0000-00-00 00":00:00 |
acl     | varchar(50)       |      |     |                     |

Now my question : Is this problem a hardware problem (are there any sun specialists here) or are there any known issues with solaris/sparc/a5000 and MySQL any special configuration needed.The server is not under heavy load compared to a single 600 MHz CPU Intel/Linux machine which can handle this tasks without any problems.

One time I dealt with a support case where bring the OS patch level on a Sun resulted in a 10 fold performance increase. That would be the first thing I would try in this case.


If you are not particularly attached to your Sun, you do get a better bang for the buck out of x86 machines, and very possibly a better bang period regardless of the buck. In 3 years of working on the MySQL support team I have never seen a Sparc that was proven to outperform a top of the line dual x86 machine.


-- Sasha Pachev Create online surveys at http://www.surveyz.com/

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]



Reply via email to