On Tue, 19 Jul 2005, Barry Murphy wrote:
[...]
>> I'm guessing customers connect directly to the erx's, thus if the erxcrash's
>> and goes down, customer falling off that node will be effected.
>>
>> The idea is a pretty map with colors indicating packet loss.
>> Let's just see what happens when the next big network fault happens.
>It is highly unlikely anyone will give you this information (and in some
>cases this is 100% unlikely) as this would tend to be considered
>"commercially sensitive" data to many large organisations.
>Any information gathered through probing networks will most certainly be
>inaccurate and could lead (as pointed out earlier) to increased
>frustration in helpdesks when some clueless customer rings up to report an
>ERX crash because they saw it on your intarwebnet page.
There's a long history of people trying to understand routing or other changes using RTT's, with names like "Internet Weather Report". The classic problem with doing this on normal (e.g. DSL) customer links is that when that customer link becomes congested, packet loss and RTT go up. But that doesn't indicate that the network as a whole is doing anything unusual.
This problem shouldn't exist for a node connected by high-speed link to (or at) an exchange point, but that approach is about looking at backbone links, not the network all the way out the the customer.
The key question is what exactly it is anyone's trying to measure.
- Donald Neal
"This communication, including any attachments, is confidential.
If you are not the intended recipient, you should not read
it - please contact me immediately, destroy it, and do not
copy or use any part of this communication or disclose
anything about it. Thank you. Please note that this
communication does not designate an information system for
the purposes of the Electronic Transactions Act 2002."
