Hi,
I actually didn't change any of the behavior options in my basic tests,
which is perhaps why we were seeing no performance improvements.
I must say that we're very excited about the consistent hashing though.
In any case, this post wasn't supposed to be about libmemcached... All I
really want to know is if someone has any idea why a basic counter using
Cache::Memcached would be failing under heavy (but not ludicrously
heavy) load...
Kind regards
-Andrew
Brian Aker wrote:
Hi!
On Apr 15, 2008, at 5:59 AM, Andrew Bryan wrote:
Well that's what I was expecting, but from my tests I was getting that
it (Cache::Memcached::libmemcached) was slightly slower (or at least
roughly the same; but certainly not an order of magnitude faster!).
What options did you use? Changing the behavior flags will modify the
behavior of the driver.
Cheers,
-Brian
