Re: [Sqlalchemy-tickets] [sqlalchemy] #2174: .7 performance regression for count queries with MyISAM

[sqlalchemy <mi...@zz...>]

#2174: .7 performance regression for count queries with MyISAM
---------------------------------+------------------------------------------
    Reporter:  guest             |       Owner:  zzzeek                
        Type:  defect            |      Status:  closed                
    Priority:  medium            |   Milestone:  0.7.1                 
   Component:  orm               |    Severity:  no triage selected yet
  Resolution:  worksforme        |    Keywords:                        
Status_field:  completed/closed  |  
---------------------------------+------------------------------------------
Changes (by zzzeek):

  * status:  new => closed
  * resolution:  => worksforme
  * status_field:  awaiting triage => completed/closed


Comment:

 yup, the decision wasn't entirely easy but the new way works in all cases
 consistently, simply - the old way did not, as new kinds of bugs kept
 occurring, more exceptions and left turns kept getting made in count(),
 all just to avoid using a simple SQL expression that meets the use case
 exactly.  So if you want to optimize a simple count for MyISAM, use
 func.count():

 {{{
 #!python
 session.query(func.count(Item.id)).scalar()
 }}}

 If MyISAM is really that stupid about a simple subquery then we should
 probably add a note to the MySQL docs regarding this.

-- 
Ticket URL: <http://www.sqlalchemy.org/trac/ticket/2174#comment:1>
sqlalchemy <http://www.sqlalchemy.org/>
The Database Toolkit for Python