Re: Performance trouble at 35M rows
Posted by:
ford
Date: November 26, 2004 11:57PM
My solution was not to use MySQL, or reduce number of rows. Reducing number of rows was not an option so I had to migrate to Oracle.
I have tried using hash instead of B-Tree for indexing. Hash was faster in indexing but query time went up (unacceptable). I think there is an implied limit around 10M rows per table for indexing. Performance degrade exponentially with number of rows (it is CPU bounded and not memory or disk). I haven't tried InnoDB and I will try that.
Even though I like MySQL for most jobs, I am reaching to a conclusion "not to use MySQL for very large tables".
Subject
Views
Written By
Posted
4694
November 05, 2004 11:47PM
2488
November 07, 2004 09:02PM
2580
November 10, 2004 06:55PM
Re: Performance trouble at 35M rows
2564
November 26, 2004 11:57PM
2469
November 19, 2004 04:49AM
2508
November 26, 2004 01:43AM
2540
November 27, 2004 09:21AM
Sorry, you can't reply to this topic. It has been closed.
Content reproduced on this site is the property of the respective copyright holders.
It is not reviewed in advance by Oracle and does not necessarily represent the opinion
of Oracle or any other party.