🚀 go-pugleaf

In article <40qs2d$5o9@news.cloud9.net>,
Thor Lancelot Simon <tls@cloud9.net> wrote:
>Of course the cache increases read performance -- if you're reading data 
>which is already in the cache.  But adding memory to the machine the disk 
>is on will have exactly the same effect, and is much cheaper, too.

No, it isn't. (Adding memory to the machine).

No I can't explain _why_ it isn't, neither do I particularly CARE _why_
this is not true.  I have DEMONSTRATED the falsehood of this statement
to myself using caching controllers with heavy jobs moving memory from
core onto card.

There are lots of much more eloquent folk than I who can give you all
these wonderful reasons why the memory does more good in core as OS
buffer cache.  Sounds great, makes good sense, presents well, doesn't
hold up in the real world.

A friend of mine I turned on to DPT caching controllers called me the
other day.
He: "You know those people that argue about OS cache vs controller
cache?"
Me: "Yea."
He: "They are full of shit!"  
Me: "Yea, I know, why say you"
He: "I just installed 4MB of cache on the DPT in ... and an formerly 
8 hour job now runs in 4!"
Me: "Yea, I know."

...
--
Keith Smith <keith@ksmith.com>                  Free Usenet News/Newsfeeds
Digital Designs                                   (910) 423-4216/7389/7391
PO Box 85                                                   V.34/V.34/V.32
Hope Mills, NC 28348-0085  ... Somewhere in the Styx of North Carolina ...