![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
At 09:40 PM 12/02/1999 -0800, Bill Manning wrote: >> My point was that suggesting that reliance >> should not be placed on the DNS has been, is, and most likely will be contrary >> to what happens in the real world. >> >> Rgds, >> -drc > >Someone who used to work here at ISI had a saying, "There are no urgent >DNS problems." I did not believe then and am finding less reason to >do so as time progresses. An interesting viewpoint is espoused in the >recent draft by Akien et.al. on Middleware. According to that draft, >we are headed into an Infrastructure that will not work w/o available >support services e.g. policy db, CA/key servers, DHCP/DNS, ad.nausea. > > >-- >--bill > The Army's efforts to use COTS Internet technology in a tactical environment has run straight into the critical support services issues. Without the functioning servers that Bill mentions, the network of applications is hung. Funny, they don't use cell phones because you have to build an infrastructure, but we keep trying to use fixed-plant technology for data networking. For the public Internet, it seems that we're building a set of critical services that, if attacked and disrupted, could crpple e-commerce, phone service, and other applications dependent on the 'Net. This came up in a discussion of "convergence" of voice, video, and data. Someone woke up to the fact that if we put all those services on one LAN infrastructure, then that LAN MUST BE UP all the time. What is the cost to keep up all the infrastructure pieces 7x24 and never let them go down? Remember, you can't even call the Help Desk if the LAN is down (assuming phone service depends on it)! The anti-convergence part of me remembers when Richmond, VA had an ice storm that took out electricity for a week last Xmas. It was very interesting to note that we still had phone service, natural gas delivery, and water pressure. Each infrastructure piece had separate delivery mechanisms and their own backup apparatus. The loss of any one had almost no effect on the others. Is this what convergence would get us? Careful what you wish for... :-) Walt
Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.