October 1996 - Test - lists.nanog.org

Re: Cisco's AIP vs HSSI
by Vadim Antonov 16 Oct '96

16 Oct '96

Kent W. England <kwe(a)6SigmaNets.com> wrote: >But I like PPP because cisco still doesn't document their HDLC line >protocol. :-) (Go ahead, try to find it online or ask someone on >one of the tech support lists to find it for you. :-) Look in BSD/386 source code (/sys/net/if_c_hdlc*). I reverse-engineered their protocol quite a time ago, took a day of looking at bytes going back and forth :) --vadim

1 0

Re: BBN outage [Press Release]
by cdc＠GroupZ.net 16 Oct '96

16 Oct '96

FYI forwarded from the NANOG group - cdc Hi, Here's the formal Press Release on the outage...complete with down times for those of you who were asking for specifics. Note - I work the gov't services side of things, so don't let the BBN mailing address fool anyone. I'm usually not privy to real time happenings on the commercial side, but I can follow up on high visibility issues when requested...like "The CIDR List" and this outage, etc. --------------Forwarded Message: Date: Fri, 11 Oct 1996 15:31:30 -0500 To: BBN-Staff(a)BBN.COM From: "Joan L. Irons" <jirons(a)BBN.COM> Subject: Power Outage Release FOR IMMEDIATE RELEASE For Information Contact: Peter Thonis BBN Corporation 617-873-3512 pthonis(a)bbn.com STANFORD POWER OUTAGE AFFECTS SOME BBN INTERNET CUSTOMERS CAMBRIDGE, Mass., October 11, 1996 - An electrical power outage at the Stanford University complex has caused BBN Corporation's Internet access in Palo Alto, California to be temporarily unavailable, affecting approximately 400 of BBN's business and organizational customers. The electrical outage, which began at 7 p.m. PDT on Thursday, has cut off power to BBN's Point of Presence (PoP) located at Stanford University Data Center. The PoP is one of the connection points to the Internet for BBN's customers in California. BBN has more than 60 PoPs in the U.S. A battery backup power system kept the BBN Palo Alto PoP operational for two hours, but power was lost at 9:00 p.m. PDT. Stanford's power was restored at 12:30 a.m. on Friday and then went out again at 6:00 a.m. PDT. Stanford has a redundant power system, tying in Stanford's co-generation facility and PG&E Power, but the equipment that switches between these power sources failed. This is the first power failure at the Stanford facility in more than 10 years. Backup generators are being deployed to the Stanford site. BBN expects Internet access through the PoP to resume later today, once electrical power is restored. BBN Chairman and CEO George H. Conrades, who is at the Palo Alto site working with customers on the problem, said, "We apologize for the fact Internet access is unavailable while Stanford University works to repair its electrical power. We understand how important full-time access to the Internet is to our customers, and we are doing everything we can to help Stanford restore electrical power to BBN's facilities. Up to this point, the facility has had an excellent history of uninterrupted power service, including through brownouts and earthquakes. Obviously, we are committed to providing our customers with the highest standard of reliable service." # # # ------------------------------------------------------------ Joan L. Irons Corporate Communications (617) 873-3514 Voice BBN Corporation (617) 873-6899 Fax 150 CambridgePark Drive jirons(a)bbn.com Internet Cambridge, MA 02140 http://www.bbn.com ---------------End Forwarded Message

1 0

Re: BBN outage [Press Release]
by Patty Martinez 16 Oct '96

16 Oct '96

Hi, Here's the formal Press Release on the outage...complete with down times for those of you who were asking for specifics. Note - I work the gov't services side of things, so don't let the BBN mailing address fool anyone. I'm usually not privy to real time happenings on the commercial side, but I can follow up on high visibility issues when requested...like "The CIDR List" and this outage, etc. --------------Forwarded Message: Date: Fri, 11 Oct 1996 15:31:30 -0500 To: BBN-Staff(a)BBN.COM From: "Joan L. Irons" <jirons(a)BBN.COM> Subject: Power Outage Release FOR IMMEDIATE RELEASE For Information Contact: Peter Thonis BBN Corporation 617-873-3512 pthonis(a)bbn.com STANFORD POWER OUTAGE AFFECTS SOME BBN INTERNET CUSTOMERS CAMBRIDGE, Mass., October 11, 1996 - An electrical power outage at the Stanford University complex has caused BBN Corporation's Internet access in Palo Alto, California to be temporarily unavailable, affecting approximately 400 of BBN's business and organizational customers. The electrical outage, which began at 7 p.m. PDT on Thursday, has cut off power to BBN's Point of Presence (PoP) located at Stanford University Data Center. The PoP is one of the connection points to the Internet for BBN's customers in California. BBN has more than 60 PoPs in the U.S. A battery backup power system kept the BBN Palo Alto PoP operational for two hours, but power was lost at 9:00 p.m. PDT. Stanford's power was restored at 12:30 a.m. on Friday and then went out again at 6:00 a.m. PDT. Stanford has a redundant power system, tying in Stanford's co-generation facility and PG&E Power, but the equipment that switches between these power sources failed. This is the first power failure at the Stanford facility in more than 10 years. Backup generators are being deployed to the Stanford site. BBN expects Internet access through the PoP to resume later today, once electrical power is restored. BBN Chairman and CEO George H. Conrades, who is at the Palo Alto site working with customers on the problem, said, "We apologize for the fact Internet access is unavailable while Stanford University works to repair its electrical power. We understand how important full-time access to the Internet is to our customers, and we are doing everything we can to help Stanford restore electrical power to BBN's facilities. Up to this point, the facility has had an excellent history of uninterrupted power service, including through brownouts and earthquakes. Obviously, we are committed to providing our customers with the highest standard of reliable service." # # # ------------------------------------------------------------ Joan L. Irons Corporate Communications (617) 873-3514 Voice BBN Corporation (617) 873-6899 Fax 150 CambridgePark Drive jirons(a)bbn.com Internet Cambridge, MA 02140 http://www.bbn.com ---------------End Forwarded Message

1 0

Re: Cisco's AIP vs HSSI
by Kent W. England 16 Oct '96

16 Oct '96

At 02:15 PM 15-10-96 -0400, Avi Freedman wrote: >> ..., IMHO, you really need PPP. > >We run PPP on all non-frame-or-smds and less-than-DS3 links so that >the customer (or ourselves, if it came to it) could switch to non- >Cisco gear instantly. > >But we usually leave most cisco-cisco high-speed links at HDLC. >My impression is that HDLC was the same efficiency - or moderately >more so - than PPP. There are about three flavors of HDLC encapsulation, one of which is cisco and because of that the others don't matter. There is precious little difference between all the flavors of HDLC and PPP, they all carry the old OUI/Ethertype/etc overhead. And the efficiencies of all are the same compared to SMDS or ATM. > >For what do feel that one *really* needs PPP? > Well, that's it -- I don't assume that the whole world is cisco, but I'm a bit naive. Of course there are other hooks in PPP that are useful, since PPP, like BGP, is a Swiss Army knife protocol, but again since the whole world is cisco, let BGP figure everything out and run HDLC. You don't really need PPP in an all cisco backbone. But I like PPP because cisco still doesn't document their HDLC line protocol. :-) (Go ahead, try to find it online or ask someone on one of the tech support lists to find it for you. :-) --Kent I'm kidding. Of course I love everything about cisco, since as Dave O'Leary has told me for a long time "Sooner or later everyone ends up working for cisco" and he seems to be right.

1 0

RAToolSet 3.4.1 is available for anonymous ftp
by Cengiz Alaettinoglu 16 Oct '96

16 Oct '96

RAToolSet 3.4.1 (contains RtConfig and Route Object Editor in addition to other tools) is now available at http://www.isi.edu/ra/RAToolSet ftp://ftp.isi.edu/ra/RAToolSet The ftp directory also contains binaries for selected architectures. We can make binaries for other systems available upon request. RAToolSet 3.4.1 compiles out of the box for the following systems: SunOS, Solaris, OSF/1, BSDI, FreeBSD, linux, and probably on other systems as well. There is also a mailing list, ratoolset(a)isi.edu. To join, please send a message to majordomo(a)isi.edu with body containing subscribe ratoolset In the upcoming Nanog, there will be demos of these tool. Bug reports, fixes and suggestions for improvements are welcome and appreciated. Cengiz -- Cengiz Alaettinoglu Information Sciences Institute (310) 822-1511 University of Southern California http://www.isi.edu/~cengiz

1 0

Parking info for NANOG Meeting
by Pam Ciesla 16 Oct '96

16 Oct '96

There is LIMITED PARKING in the Thompson Parking Structure (ONLY) located on Thompson St. about 1 block from the MI Union. Please park your car on the 2nd level and up (do not use the 1st level). You will need to come into the meeting site to get a parking pass and return to your car to put it in your car. Please NOTE that if you get a ticket while doing this bring it to me and I will take care of it. PARK IN STAFF PARKING. NOTE: THIS IS A FIRST COME FIRST SERVE POLICY. HOTEL SHUTTLES ARE HIGHLY RECOMMENDED. Pam Ciesla ps There are no meters available in the Thompson Parking Structure, please use Staff Parking spaces. Sorry for any confusion.

1 0

NANOG Registration
by Pam Ciesla 16 Oct '96

16 Oct '96

The list of attendees for the NANOG Meeting has been updated. If you are not on the list and should be please contact me. NOTE: THIS IS LAST CALL FOR REGISTRATION FOR THE OCTOBER NANOG MEETING Hotels that were blocked are no longer available. Please check other hotels list on the WEB page also check the Crowne Plaza 313-761-7800. If you need any info please contact me. Thanks Pam Ciesla

1 0

Re: Cisco's AIP vs HSSI
by salo＠msc.edu 16 Oct '96

16 Oct '96

> From: bmanning(a)ISI.EDU (Bill Manning) > Subject: Re: Cisco's AIP vs HSSI > To: freedman(a)netaxs.com (Avi Freedman) > Date: Tue, 15 Oct 1996 07:55:50 -0700 (PDT) > Cc: bmanning(a)ISI.EDU, salo(a)msc.edu, kri(a)bellcore.com, nanog(a)merit.edu > [...] > The orginal question can be paraphrased as; "which > board is better" (see the subject line). > Tim S. couched his replies in the form of an ATM only > solution. I feel that the HSSI board was not designed > as such. Your comments seem to back up my assertions. The HSSI card was not designed as an ATM solution. The HSSI ATM solution involved, in some sense, making the ATM network look like a frame relay network to the HSSI card. > I would claim that for raw throughput, and assuming > the DS3 daughtercard for the AIP, that the HSSI board > delivers more useful bits at the expense of an outboard > CSU. This is based on my feeling that HDLC is more efficent > than ATM signaling/framing. If the context of the question is the most appropriate interface for a point-to-point connection between two routers, I would worry more about throughput than, for example, issues of ATM versus HDLC functionality and overhead. The AIP card can obviously clock out data a lot faster than can a HSSI board, (assuming you don't intentionally put a slow interface on the AIP card). However, I think it is also important to understand the throughput of the two boards and how well they are integrated into the Cisco (or, in general, your favorite vendor's) router. I was under an impression that the HSSI card was a lower-performing board, but someone with some more detailed understanding should provide some insight. I suppose my answer could be refined as: o If you need an ATM solution, use the AIP card. o If you merely want a DS-3 or slower point-to-point link between two routers, test both solutions, (or talk to someone who has), to determine which solution has better performance; or o If you want a point-to-point connection between two routers at speeds faster than HSSI can support (a bit more than 45 Mbps, I think) use the AIP card. Again, you should also consult your router vendor. They should be giving you all this information and more. For example, you might discuss a OC-3c IP over SONET solution for point-to-point connections. -tjs

1 0

Re: cable modems [Re: Intra/Inter - was Inet-II]
by Kent W. England 16 Oct '96

16 Oct '96

At 09:01 AM 15-10-96 -0700, Milo Medin wrote: >Really Ken, just because you consult to PacBell, you shouldn't be so >obvious in your posturing. :-) Well, Milow, it's hardly posturing. More a gesture. :-) Everyone who's been to NANOG knows that I work for PacBell on the NAP. I believe that you yourself were once a consultant for a long stint at Ames, working for Sterling Software before you stepped over and became a government employee? I don't think you can so easily discredit me for being a low-life self-prostituting money-grubbing consultant. But nice try. I'm sure you've read the Upside article and the issue it raises regarding TCI's investment in the necessary cable plant upgrades to support @Home? > >Cable modems by themselves don't don't the performance problem, in fact, >that's what @Home has been talking about for sometime now, and why we >believe you have to build an intelligent network architecture, not just >dumb pipes. That is certainly true and no one better to architect a data network that works for the web than you. An xDSL network needs the same kind of caching and replication as you are designing into the @Home network for the same reasons. Fortunately, it isn't rocket science (to take your old NASA phrase) and doesn't require proprietary extensions, although some might like to lock in a subscriber to proprietary features of an intelligent web architecture. > >Some folks do believe that just fixing the loop will fix system >performance, and they are clearly in the wrong. However, I will venture >to say that unless you do something like what we're doing, it doesn't >matter if the loop is ADSL or Cable or laser, it's not going to end up >running very fast. This is not just a loop problem. > Right again, just fixing the loop problem isn't enough, but since we are tangentially back on that topic, what exactly has TCI and the other cable partners committed to regarding the investment in cable plants outside the Bay Area? For while you absolutely must have an intelligent data architecture, you also absolutely must have a solution to the local loop problem. You go fix the local loops at TCI and I'll get PacBell to roll out an intelligent data architecture along with the xDSL service. --Kent

1 0

Re: Cisco's AIP vs HSSI
by Kent W. England 16 Oct '96

16 Oct '96

At 10:58 AM 15-10-96 -0400, Avi Freedman wrote: >> This is based on my feeling that HDLC is more efficent >> than ATM signaling/framing. > >Agreed... >But it's not a feeling, it's a fact :) > >> --bill > >Avi > > > While this is certainly true, HDLC is a point-to-point protocol and not a network protocol, like frame relay, SMDS or ATM. And HDLC itself isn't quite enough, IMHO, you really need PPP. And the efficiency lost to ATM is not 40% as often claimed on this list, but rather it is 12% less efficient than PPP for TCP. 10% is the cell header overhead and 2% is due to modulo 48 padding, given actual traffic at FIX West as measured by kc at NLANR. --Kent We return to our regularly scheduled ATM tweaking program now in progress. :-)

2 1