[MTOS-dev] Performance

[Alvar Freude]

Hi,

-- Hirotaka Ogawa <hirotaka.ogawa at gmail.com> wrote:

> So, after applying attached patch, can you please run a small benchmark
> as like:

only a quick&dirty test:

With really small test DB, and removed ->ping call (line 68 in 
Data::ObjectDriver::Driver::DBI):

PostgreSQL:
MT::Entry->exist: 1.348955
MT::Entry->count: 1.427626

MySQL (and nearly empty database!):
MT::Entry->exist: 1.473078
MT::Entry->count: 1.496623

(no stable values; perhaps I may test it with the Benchmark Module the 
next days)


With the ->ping, PostgreSQL:

MT::Entry->exist: 1.540907
MT::Entry->count: 1.639969


MySQL:

MT::Entry->exist: 1.632844
MT::Entry->count: 1.66519


> If 'exist' is faster than 'count', it is worth the extra effort to
> rewrite MT to use it.

it is faster, but the time is not very exact and varying.
And it is only about 10%, because of the Perl overhead.

My Testing tables are very small (1 blog for MySQL, 6 for PostgreSQL ;) 
)! With bigger Database the difference may grow.


For MySQL, it's always the count, no exist; with the exist for MySQL the 
performance is the same.


The ->ping costs for both nearly 20%.


Ciao
  Alvar

-- 
** Alvar C.H. Freude, http://alvar.a-blast.org/
** http://www.assoziations-blaster.de/
** http://www.wen-waehlen.de/
** http://odem.org/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 194 bytes
Desc: not available
Url : http://www.sixapart.com/pipermail/mtos-dev/attachments/20080306/919dfb98/attachment.bin