MySQL Forums
Forum List  »  InnoDB

Re: Select count very slow
Posted by: Peter Brawley
Date: May 23, 2017 02:24PM

The engine expects to have to examine 22.5 million rows, that 10% will be removed by the treated=0 condition, and that no available index will speed things up.

This takes 10 mins on machine 1, 2 mins on machine 2? What besides OS are the diffs between the two machines?

If the average row is 250 bytes, that's 5GB of indexes & data. You can check that number with ...

select   
  concat(round(sum(table_rows)/1000000,2),'m') rows,  
  concat(round(sum(data_length)/(1024*1024*1024),2),'g') data,  
  concat(round(sum(index_length)/(1024*1024*1024),2),'g') idx,  
  concat(round(sum(data_length+index_length)/(1024*1024*1024),2),'g') total_size,  
  round(sum(index_length)/sum(data_length+index_length),2) idxfrac  
from information_schema.tables  
where table_schema=            -- FILL IN YOUR DB NAME
  and table_name='record'; 
[/code[

It does take a long time to read and return so much. But it's hard to imagine the MySQL engine taking 10 mins to just read 22 million rows unless there's a lot of  swapping going on. It's also hard to imagine a reason to return all 22 million rows need in one go. Can you clarify?

If you disagree with the engine's estimates as given by Explain, you could lay in an index (id,treated), tell the engine to use it, and see if it's quicker.



Edited 1 time(s). Last edit at 05/23/2017 02:24PM by Peter Brawley.

Options: ReplyQuote


Subject
Views
Written By
Posted
372
May 22, 2017 09:32AM
209
May 22, 2017 11:24AM
170
May 22, 2017 11:38AM
153
May 23, 2017 10:43AM
195
May 23, 2017 11:10AM
Re: Select count very slow
139
May 23, 2017 02:24PM
159
May 23, 2017 04:05PM
138
May 23, 2017 08:04PM


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.