EUGENE@vm.utcs.utoronto.ca (Eugene Siciunas) (01/11/91)
A question has been asked about current CA*net status ..... Since most of the info flow about CA*net never made it onto the NRCNET-L list, which this list subsumed, here goes.... As of today, Jan 11, all Canadian regional networks are interconnected by the CA*net backbone with 56 Kbps links. There exist three links between CA*net and the US NSFNET, but only one of the three, from Montreal to Princeton, could be considered "in production", and it is the planned 112 Kbps link. A 56 Kbps link exists between SFU in Vancouver and the U of Washington in Seattle, but due to rerquired upgrades to the regional Proteon router at SFU, no traffic is as yet being routed through that link. Instead, US bound traffic from the BC or Alta regionals hits the US through the Toronto link to Cornell. This situation will hopefully be resolved soon. The Toronto to Cornell connection is presently effected through two separate 56 Kbps links. The plans called for a 168 Kbps (3x56) link, but this proved to be technically unprovideable, so a 224 Kbps (4x56) is almost ready. It is expected any day now.... So CA*net is up and running, (and has been with coast-to-coast connectivity) since early October. At that time, not all domestic links were 56 Kbps, but they now are. The initial topology emulated two overlapping, not quite closed loops. With the tariff changes (reductions) announced by the CRTC last Dec 1, we have ordered more lines to "close the loops", specifically between Edmonton and Saskatoon, and St. John's and Charlottetown. This will provide redundancy, in that if one path to a node becomes disabled, it will still be reachable, (automatically) by data flowing to it from the other direction. This is expected to happen as soon as the new links are installed, hopefully before the end of January. Further topology/link speed changes can be made as $$ become available, either through reduced tariffs or influx of funds, and traffic patterns warrant.
pkern@utcs.utoronto.ca (Paul Kern) (01/11/91)
In article <91Jan11.091912est.18503@gpu.utcs.utoronto.ca> EUGENE@vm.utcs.utoronto.ca (Eugene Siciunas) writes: >The initial topology emulated two overlapping, not quite closed loops. >With the tariff changes (reductions) announced by the CRTC last Dec 1, >we have ordered more lines to "close the loops", specifically between >Edmonton and Saskatoon, and St. John's and Charlottetown. This will >provide redundancy, in that if one path to a node becomes disabled, >it will still be reachable, (automatically) by data flowing to it >from the other direction. This is expected to happen as soon as the >new links are installed, hopefully before the end of January. > For those completely unfamiliar with CA*net (ie. some of those subscribed to this newsgroup) the current physical topology looks something like ... Edmonton ...... Saskatoon | | | | NSFNet@Seattle ------- Vancouver Winnipeg | | | | NSFNet@Cornell ======= Toronto ------- Montreal ====== NSFNet@Princeton | | | | Fredricton Halifax | | | | Charlottetown .... St. John's line legend: --- 56 kbps === 112 kbps ... "under contruction"
prindevi@inria.inria.fr (Philippe-Andre Prindeville) (01/11/91)
Just out of curiousity, what is the EGP you are running, and how extensively instrumented is the network? What version of MIB are you using, and with which extensions? If you are using BGP, are you using the proposed BGP MIB? Can the regionals collect traffic stats from the routers? What about event monitoring? Thanks, -Philip