GDA interim benchmarking synopsis

[Derek Atkins]

Mark Johnson <mrj001 at shaw.ca> writes:

>> I'm really beginning to wonder if we need to write special SQL
>> based on the database we're using in order to tune the performance
>> because of issues like this?  Yes, I know that that sort of defeats
>> the purpose of using a library like GDA, but if GDA can't form
>> an optimal query then it might behoove us to do the right thing
>> rather than hope they do.
>>   
> I hope not, but we are talking not about minor differences in
> performance, but orders of magnitude.  I was absolutely floored at
> MySql's performance with sub-queries.  I do have a couple of things to
> try with it, though:
> 1. creation of a secondary index on slots.obj_guid
> 2. There is a section of the MySql manual on optimizing this stuff.
> Maybe it has something helpful.

Exactly.  If it were just a minor difference then I just wouldn't
care, but the difference between seconds and minutes is sufficiently
large that we may need to work around the issue ourselves.

-derek
-- 
       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord at MIT.EDU                        PGP key available