mailing list archives
Real-life figures of hardware limitations
From: Stefan Baltus <stefan.baltus () xbn nl>
Date: Wed, 6 Aug 2003 22:16:28 +0200
As network-specialists often design new networks and evaluate
equipment of different vendors I can imagine that, in my own
finding, more people find it hard to determine the limitations of
what equipment can do in terms of maximum throughput of a
certain type of interface under certain conditions, the maximum
number of accounting data (such as flow-statistics) a router
can really sustain, how much ATM vc's you can really terminate
on a type-x router that carries x Mbit/s of traffic, etc.
Most of the time, the technical specs are either over-promising,
have 'sales&marketing'-type of figures or do not really apply
to real-life traffic patterns. Also, it is not always possible to
do real-life testing with the box, due to a number of
problems that can arise (can't get hold of the box, can't
generate enough traffic, can't generate traffic that is a
good representation of the real-life situation, etc).
I am looking for information that gives me (and, if people are
willing to share, I'm willing to put all data on a publicly
available website) more insight in these kind of figures. It is
possible that this is already being provided on a website of
somebody, but I haven't found it yet; people that have, please
point me in the right direction.
The goal is to assemble as much information as possible to
provide the myself and everyone who wants to know with real-life
information about what a certain piece of hardware can do, to
complement and/or confirm the figures from the spec-sheet.
Please reply off-list. I will post a summary on the list
for the people that are interested.
Stefan Baltus <stefan.baltus () xbn nl> XB Networks B.V.
Manager Engineering Televisieweg 2
telefoon: +31 36 5462400 1322 AC Almere
fax: +31 36 5462424 The Netherlands
- Real-life figures of hardware limitations Stefan Baltus (Aug 06)