We currently have a MySQL 5.1 production server running on CentOS 5.5. The database has about 5 millions rows. We have a fairly complex query used to generate a report which involves multiple table joins. Its fully indexed and optimized. On our production server, it runs consistently in about .4 ms.
We have built a new system with twice the ram running on CentOS 5.5 and MySQL 5.5.
I've tweaked the config file on the new machine with similar settings as our current production machine, adjusting some things to take advantage of the additional memory. I did a mysqldump of the current production database system and the imported it into the new 5.5 system.
The problem is the same query which runs on the production system in ~.4 ms take ~2.5 seconds on the new machine. Even worse, re-running it over and over consistently adds ~5 to the run time (ie 2.5, 8, 13.5 seconds ...). The tables and indexes are identical.
I've never seen this issue and have no idea what to look for which could be causing this. Let me know what relevant information I can post in helping to solve this.
Thanks,
Anthony
We have built a new system with twice the ram running on CentOS 5.5 and MySQL 5.5.
I've tweaked the config file on the new machine with similar settings as our current production machine, adjusting some things to take advantage of the additional memory. I did a mysqldump of the current production database system and the imported it into the new 5.5 system.
The problem is the same query which runs on the production system in ~.4 ms take ~2.5 seconds on the new machine. Even worse, re-running it over and over consistently adds ~5 to the run time (ie 2.5, 8, 13.5 seconds ...). The tables and indexes are identical.
I've never seen this issue and have no idea what to look for which could be causing this. Let me know what relevant information I can post in helping to solve this.
Thanks,
Anthony
/pic9772200.png)
