From an operational perspective, I think the best lesson to be learned is to grow only as fast as you can provide support for.
I wonder what the correct ratio is in terms of number of new subscribers vs number of support staff. Also what is the bandwidth that they are adding from/to the CO to serve all these new customers. I hope that they are not using the good old Erlang B models to calculate the amount of backbone bandwidth they need. Bora ----- Original Message ----- From: Chris Williams <chris.williams@third-rail.net> To: Joseph Birthisel <josephb@onramp.net> Cc: <nanog@merit.edu> Sent: Monday, June 05, 2000 1:07 PM Subject: Re: DSL (was shopping for NOCs)
Doesn't this have more to do with the philosophy of the company providing service than the technology being deployed (DSL)? Maybe the problem is that DSL tends to involve LEC people who are infected with the obsolete lazy-monopoly attitude of the incumbents.
In any case, are there unique operational challenges posed by DSL that are causing some of these problems? Is there anything we can do as a community to address them (or encourage the people who need to address them, to do so)?
Maybe this discussion could evolve into something a little more useful than "my DSL provider sucks"..."me too", since there seems to be a lot of interest (or at least a lot of unhappy voices)..
Joseph Birthisel wrote:
DSL NOCs to my experience are badly run and resolution time is not good (to put it nicely). They do not keep good documentation (or their people don't choose to use it) on what circuits affect which customers, which ISPs their outages affect, or how to match a CID with a dslam. Most notable in this area is Northpoint's NCC. My hope is that either (a) over time they will get more standardized in their outage notification and communication with the ISPs they partner with (b) the technology becomes outdated soon and customer connectivity does not rely on their technical support.