Testing memcache vs db

[Jamie McCarthy]

davidb999 at gmail.com (David B) writes:

> Curious if any of you have some metrics of using memcache with a
> db…performance numbers before and after?

I think this is in the docs somewhere, but:  memcached's big
advantage is not increasing speed of the client application
directly, it is taking load off the database.  It is often very
difficult and expensive to make databases run faster.  But a chunk
of RAM and some spare CPU cycles for memcached are often cheap.

So performance will totally depend on the application, and figures
are not portable from one application to another.  In many cases,
if memcached gets your DB load down underneath the point where it
would bog queries, the "performance" increase may be infinite:  the
difference between the application responding, and not!
-- 
  Jamie McCarthy
 http://mccarthy.vg/
  jamie at mccarthy.vg