PCBUILD Archives

Personal Computer Hardware discussion List

PCBUILD@LISTSERV.ICORS.ORG

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Drew Dunn <[log in to unmask]>
Reply To:
PCBUILD - Personal Computer Hardware discussion List <[log in to unmask]>
Date:
Wed, 11 Nov 1998 16:11:18 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (30 lines)
An ISP has several sites that get their Internet feed from them.  To avoid
expensive bandwidth issues, they desire to cache web requests to those
sites.  The problem is that we cannot determine exactly what CPU to put in
these systems.

Their existing caching server has (effectively) 3 T1's connected to it.  It
is currently at capacity.  The system will become processor bound if any
more T1's are connected.  It has a PII-350 with 512MB of RAM.

The question at hand is, what processor would effectively handle four T1
connections?  Would a single PII-450 be effective or will a dual PII-350
work?  My calculations show that a PII-450 is roughly 23% faster than a
PII-350.  If we represent the pooled T1 lines as a total bandwidth of 100, I
would need a processor that can handle a bandwidth of 133.33, but I show
that a PII-450 would handle about 123.  Is that true?  Should it be more?
Less?  If the caching software forks processes to multiple processors, would
a dual PII-350 be more or less effective than a single PII-450?

One concern is that we deliver the best bang for the buck.

Thanks,

Drew Dunn
[log in to unmask]
http://adsl24.bois.uswest.net/drew/index.htm

                                  -----
                PCBUILD mailing list -  http://nospin.com
         Bob Wright:[log in to unmask] - Drew Dunn:[log in to unmask]

ATOM RSS1 RSS2