lml, thanks for the vote of confidence, but my post was an educated stab in the dark, expressing what I thought made sense. However, I am open to comments and corrections, as always, if any part of what I've stated requires it. If AOL and the ILEC can come to terms as I've described, then I think it makes a lot of sense as to why AOL's fare would be considerably less than the individual subscriber's going rate.
Re: Covad/Concentric, it was my understanding that Concentric was one of the first to offer QoS on its VPNs. If the Covad feed leverages this capability on the Concentric backbone, and a VPN is in fact being subscribed to, then I don't understand the degradation that JMD is experiencing during peak periods. Unless JMD's organization has elected to subscribe to a lower tier of service? (JMD, only you can answer this.) Frank_C.
==========
ps - I just read JMD's reply during this edit period.
>>We have NOT contracted with either Concentric or Covad to provide the VPN other than in the sense of buying a small bundle of IP addresses from Concentric for a fairly nominal annual charge. We have then "created" a VPN by way of implementing it with servers, NT, etc. This involves something called "point to point tunneling network" or PPTN which essentially encapsulates transmissions between remote workstations and the server giving us a 'secure' tunnel through the cloud, or at least that's the representation.<<
That answers my question re the VPN and the QoS. While you may be able to induce a form of tunneled environment on an end to end basis [I think?], without the backbone provider's involvement and cooperation (read: configuration and compensation), you cannot at this time command QoS. Therfore, the erratic bahavior. The subscriber's equipment (premises routers, switches and bridges) must interwork to this end with the ISP's router configurations. If you haven't subscribed to the higher tier of service, then it is highly unlikely that you will get any kind of predictable QoS. |
