MySQL Forums
Forum List  »  Newbie

Re: HANDLER statement questions
Posted by: Nikos Sarkas
Date: December 06, 2010 01:24PM

Hi Rick,

Thank you for your time. I understand what you mean and why it would impact a HANDLER READ NEXT/PREV sequence of statements. I believed that the HANDLER READ index=(value1,value2,...) variant would always reposition the cursor in the BTREE to the position specified by (value1,value2,...) and then read. If this is true, I cannot see how freshly inserted rows could be missed. But I have observed that rarely, rows do get missed (the example that I provided at my first post).

With respect to your benchmark question, Rather than substituting a large select with multiple handler reads, I was trying to substitute many selects with many handler reads, since my selects where of the "this is the index key, get me the value" variety. My benchmark was very informal and I focused on the effect on my total processing time, which decreased by 20-30%. So this is the lower bound on the improvement (query cache was disabled in both cases)

Best,
Nikos

Options: ReplyQuote


Subject
Written By
Posted
December 04, 2010 09:37PM
December 06, 2010 12:32AM
December 06, 2010 07:47AM
December 06, 2010 11:05AM
Re: HANDLER statement questions
December 06, 2010 01:24PM
December 06, 2010 10:46PM


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.