[Mimedefang] Re: md_check_against_smtp_server

[Les Mikesell]

Daniel Aquino wrote:
>> This seems like it would be a great use for memcached:
>> http://www.danga.com/memcached/
>> if you have enough traffic to make it worthwhile.  The same for any
>> other tests that are likely to be repeated often where the results could
>> be reused.  And it has the advantage of easily letting multiple machines
>> share the cache.
> 
> Yes it does !
> 
> My company only really has about 60 or so addresses so I'm definitely
> going to look into using this... I'm sure the memory needed will be
> small and it will make things run faster..
> 
> It's almost as if the pool of processes get to share data in memory 
> correct?

It's not quite as fast as direct memory access because you have the 
overhead of going through a socket, but otherwise yes.  Anywhere you 
would do some expensive query or lookup you check the cache first and if 
that fails, do it the hard way and store the result in the cache for 
next time.  The server and client library code takes care of the rest.

-- 
   Les Mikesell
    les at futuresource.com