Jump to content


Photo

Mysql performance concern


  • Please log in to reply
No replies to this topic

#1 hvle

hvle
  • Members
  • PipPipPip
  • Advanced Member
  • 667 posts
  • Locationmelbourne, Australia

Posted 04 October 2005 - 11:13 AM

I have a table name prob.
prob has 2 fields:
- ID: Interger, AutoIncrement
- Code: varchar 20

both fields are indexed and table is myISAM.

I currently have about 3 millions records in the table. In other word, I have 3 millions codes, these codes are distinct.

when I query: select * from prob where code='84l4jf983jf938'; (this code is not existed in database)

After about 6 seconds, an empty result is returned.

This post is not about error, I have no error. I only concern about the performance: 6 seconds is disasterous!!!

I suppose the indexes are store in binary tree, 3 millions records will average up to about depth 13. If it is 13 depth, how can it take 6 seconds to compare 13 20-chars string?

Any mysql expert might explain why it should take ridiculously long, and how can I optimize it?


Life's too short for arguing.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users