- NANOG - lists.nanog.org

RE: Sprint's New ION products
by Doug Stanfield 04 Jun '98

04 Jun '98

"100 times faster" is also the language used by the cable companies and vendors of MCNS cable modems to describe their products. $200 price point and retail availability are also part of that business model. The "up to" might mean cable modems are one of a number of alternatives being considered for the last mile. Cisco has made a significant and apparently successful jump into the cable modem business so I suspect that's at least part of the plan. Doug Stanfield Oceanic Cable Data Networking Manager 200 Akamainui St. dougs(a)oceanic.com Mililani, HI 96789 >"Up to 100 times faster than conventional modems" is said in their >press release. >Perhaps this $200 unit is the NT1-equivalent. If so, it's damn cheap. >If not, why isnt billing built into the customer-end equipment? A $200 ADSL modem sounds about right and the billing will most likely be in the router/switch you are connected to, not prem-based.

4 4

Re: PPP over Ethernet?
by owen＠DeLong.SJ.CA.US 04 Jun '98

04 Jun '98

Do that with FDX Ethernet, no need to add PPP overhead to it. Owen > I guess the poster wants to connect two ethernet cards directly to > each other using a crossover cable. This could save the ethernet > overhead. Plus it would probably make for easier bonding of multiple > paralell links. Could be interesting... > > Dirk > > On Wed, Jun 03, 1998 at 11:40:22PM -0400, Perry E. Metzger wrote: > > > > Zachary DeAquila writes: > > > Anyone ever heard anything about PPP over Ethernet? Is there an > > > encapsulation standard for it? > > > > The idea sounds highly bogus. PPP is a way of adding packetization and > > content types to a serial link. Ethernet has both already. Besides, > > PPP is a point to point thing, an ethernet is a bus. > > > > .pm >

1 0

Re: different thinking on exchanging traffic
by Richard Mataka 04 Jun '98

04 Jun '98

As an interested observer who created NY IIX in New York I can tell you that getting a NAP or peering point off the ground is no easy task. One of the biggest obstacles that I have seen (IMHO) is education. There are just soooo many ISPs around who do not understand what peering is or what advantage peering can be for them!! The concept of "if you build it will they come" works in some cases but not all. With NY IIX the original idea was for it to be an international exchange point but then even regional backbone providers wanted to peer. Now we have all mixes from international to regional and many private peers. However even with close to 40 internet companies all at one location, only about 10 are currently taking advantage of the NY IIX NAP. Many of the rest are doing private peering. I know this will increase but its a wait and see attitude. I think that as the net is maturing, there are definite needs for the local NAPs, regional NAPs and then National NAPs. Working on a tiered approach may be what is needed to alleviate overloads at the MAEs and evenly distribute traffic. However, as I stated above, for this to work, education is the key as the local ISPs need to keep traffic local and etc. etc. etc... I hope that everyone doesn't mind my 1 cents worth on this thread!!!!! Regards Richard Mataka

1 0

None
by Sean Butler 04 Jun '98

04 Jun '98

John Fraizer wrote: >The thing that makes it "interesting" is the fact that most implementations >DO send an ICMP unreach back. The ICMP Unreach traffic alone generated in >the neighborhood of 1.7Mb before they routed the netblock in question to a >loopback interface on the 7507. The attacker was sending less that 300Kb >of traffic and consuming 2Mb. Any idea where that much amplification is coming from? For smurf with an echo request to a broadcast, its easy to see why there is so much amplification. But for a TCP or UDP packet to port 0, wouldn't just one port unreachable be sent back to the (spoofed) source? Or is it a broadcast TCP or UDP packet to port 0 ??? Thanks, Sean Butler, IBM Global Services

1 0

US West and RADSL
by Marianne Granoff 04 Jun '98

04 Jun '98

>On Wed, 3 Jun 1998, Jonathan Arneault wrote: > >> I just read (http://www.internews.com/isp-news/1998/06/0301-bell.html) that >> Bell Atlantic is looking to finally roll out xDSL. However, not as a >> wholesale, unbundled service, but as an Internet service which will directly >> compete against ISPs. >[snip] >> for us "little guys" to compete against the behemoths? Has anyone had >> similar experiences with their primary iLEC (such as PACBell and US West)? > >US West just rolled out RADSL in Utah. > >$40/month for a 256k bi-directional DSL line > >USWEST.NET, US West's ISP, will sell you dedicated internet access over >that connection for $19.95/month. > >Of course, the number to order either service is the same. > >Of course when you call to order the DSL line they might give you a hard >sell to also sign-up with USWEST.NET. > >Of course they might give out inconsistent and false information that >changes with every time you talk to them. > >We are a DSL enabled ISP. Our first DSL customer ordered his DSL line and >told them he wanted us as his ISP. When the installed his line, they >mapped him to USWEST.NET's ISP!!! I also know of two other confirmed >cases were the customer chose a local DSL ISP, and they mapped them >instead to USWEST.NET. > >I could go on and on about the VERY APPALLING situation here with USW and >DSL and Internet access, about regulated and unregulated services, etc, >etc. > >Dax Kelson >Internet Connect, Inc. The preceeding message was forwarded to me by one of our NM ISPs. US West has been co-marketing its Internet service _with_ its RADSL service in all of its 14 states (http://www.uswest.com/com/customers/interprise/dsl/) Actions by ISP groups in Oregon and New Mexico may provide some relief to local ISPs there. I have just put up a listserve for ISPs in the US West states to use in sharing information. Technet has had one for the NM ISPs for several years - it has been a big help in getting out the word about some of these actions. How about uswisp(a)lists.nm.net? Please feel free to send this out to any interested ISPs. To subscribe, just send an empty note to: uswisp-subscribe(a)lists.nm.org I believe that all the RBOCs/ILECs have taken or will be taking similar actions. Even Sprint's new ION services are part of this trend. In my opinion, this is anti-competitive behavior by monopoly organizations. I think that many local ISPs will be severly hurt by such actions, and more than a few will close their doors. As I see it - the biggest problem is that local ISPs are not organized and do not know how - or have the forums - to work together to fight actions by a company the size of US West (or other RBOC/ILEC). My company, New Mexico Technet, is one of the larger ISPs in NM. We wholesale Internet access to other ISPs. We have intervened in the NM tariff filing for US West's Megabit services (see http://www.technet.nm.org/press.htm) to attempt to correct some of the things that are very anti-competitive about the proposed tariff. So far it has cost us over $30,000 in legal fees and we have not even had the hearing yet. Most local ISPs cannot do this. Most local ISPs do not know how to take the actions with the FCC or with the state public regulatory agencies so that their concerns can even be heard. Frankly - most ISPs are not members of CIX or of ISP/C - and many of them do not even know about those organizations, or understand why they should care. In NM, the local ISPs come in mostly 2 flavors: those that serve urban areas (Albuquerque, Santa Fe, Las Cruces, Los Alamos) who "may" be affected by what US West does but are not sure, and those in rural areas who do not feel they will ever be affected by these actions, and so do not care. In much of US West's 14 state territory - I suspect that this is similar. The local ISPs in the urban areas have mostly seen the other local ISPs as competitors, not collaborators. They have seen US West (or the RBOC/ILEC) as a vendor, not as a competitor. Most local ISPs worry about retribution from US West (or other RBOC/ILEC) (delayed service, unresponsiveness on outages, unfilled orders) if they come on too strong in criticizing a company that they are _so_ dependent on. I am not sure that their concerns are not valid. The local (state) regulatory agencies are overworked and underfunded in this age of telecommunications transition/revolution. It is not that they don't care. It is that they simply have too much on their plates already. Few states provided extra funding to handle all the _new_ issues raised by the 1996 Telecommunications Act at the _state_ regulatory agency level. No one is championing any of this in most states. I think probably because it is not considered a problem by the vast majority of ISP _customers_. I had one of my customers tell me to "just get out of the way" and let US West introduce the high speed service because the customer needed it right away and I was just holding it up. They never saw that US West owned some of the blame in the constant delays, counterfilings, interrogatories, motions to compel, and other actions that have caused this intervention to drag on. Unfortunately, this person is more typical of ISP customers than local ISPs want to believe. The saddest aspect of this is that unless something changes, US West and the other RBOCs/ILECs will likely dominate the supply of Internet access in large urban areas in a few years - and the rural areas will have a great deal less Internet access than they do today. I think the answer is that the local ISPs _and their customers_ have to come together if they want to have choices about ISPs in the future. It will take some of the larger ISPs reaching out to the smaller local ISPs to help them get _all_ of their respective customers informed of the issues. It will take the larger ISPs intervening in more state and FCC proceedings. It will take constant email, listserves, and newsgroups spreading such information - and reaching customers - not just ISPs. It will take some national politicians to "champion" this cause - some who are not worried about losing RBOC/ILEC campaign contributions - which are considerable. It will take involvement by media organizations that are not worried about losing the RBOC advertising revenues - which are also considerable. It will take every local ISP who is harmed by US West actions calling, writing, or emailing their local and national politicians and letting them know that they have informed all of their own customers about the actions by US West or other RBOC/ILEC and informing their customers of the fact that the politician has not responded to these illegal/unethical/anti-competitive actions. In NM, it is now other internet professionals and businesses that have joined ISPs in questioning the actions of US West. Web designers, web page hosting services, internet trainers, web-advertising services and other businesses are starting to realize that US West wants to take their Internet-based business as well. This is a start. Regards, Marianne Marianne Granoff Director of Operations New Mexico Technet, Inc. 5921 Jefferson NE Albuquerque, N.M. 87109 Ph: (505) 345-6555 FAX: (505) 345-6559 email: granoff(a)nm.net or granoff(a)technet.nm.org

1 0

Re: different thinking on exchanging traffic
by shields＠crosslink.net 04 Jun '98

04 Jun '98

In article <Pine.LNX.3.95.980604003057.723W-100000(a)tarkin.fdt.net>, Jon Lewis <jlewis(a)inorganic5.fdt.net> wrote: > I went one step further and talked to some of the local ISPs about pooling > some $ and having a NAP-owned nnrp server. I don't understand the value of having it be NAP-owned vs. being owned by a third party and located at the NAP. -- Shields, CrossLink.

1 0

ADSL & iLEC
by Jonathan Arneault 04 Jun '98

04 Jun '98

I just read (http://www.internews.com/isp-news/1998/06/0301-bell.html) that Bell Atlantic is looking to finally roll out xDSL. However, not as a wholesale, unbundled service, but as an Internet service which will directly compete against ISPs. While I do believe that BA and the rest of the iLECs have the right to compete in their market, I don't think they should have the right to put private, unsubsidized businesses out of business by using their utility based advantages. Additionally, how is it that a "regulated monopoly" is able to transit service outside of the LATA (such as is required with Internet service)? IS this Bell's way of getting into LD, through packet switching? Does anyone know, now that BA has announced that they will provide ADSL internet services, if they provide a similar, tarriffed, unbundled xDSL service by which a local, regional or national ISP could provide a competitively priced service? How does a ten million dollar a year company serve in a market with a 10 billion dollar a year company? Is there a way for us "little guys" to compete against the behemoths? Has anyone had similar experiences with their primary iLEC (such as PACBell and US West)? Hoping for good news, Jonathan Arneault Director of Business Development CMA / INet Solutions Division 518.783.9003 x 253

4 3

Re: Attack/DoS
by John Fraizer 04 Jun '98

04 Jun '98

At 11:38 PM 6/3/98 -0400, Perry E. Metzger wrote: > >"Todd R. Stroup" writes: >> Don't know if it is just me. But over the last 10 hours we have been >> seeing attacks on port 0 from port 0 (both tcp and udp) on several clients >> networks. I have also seen the same attack on port udp 53(DNS). >> >> Anyone have any information on this? > >What do you mean by an "attack"? Are you being flooded? Are the >packets somehow "interesting"? Without details the information is >useless. > >Port 0, btw, is not generally valid, and most proper TCP and UDP >implementations will just send an ICMP Unreachable back when they get >such a packet. > >Perry Perry, Here are some logs of slightly different format that show the same attack Todd writes about: Jun 3 17:02:47 eth0-core0 kernel: IP acct in eth0 UDP 199.199.125.28:53 209.115.17.67:53 L=57 S=0x00 I=47916 F=0x0000 T=49 Jun 3 17:02:47 eth0-core0 kernel: IP acct out eth2 UDP 199.199.125.28:53 209.115.17.67:53 L=57 S=0x00 I=47916 F=0x0000 T=48 Jun 3 17:02:47 eth0-core0 kernel: IP acct out eth0 ICMP/3 209.115.17.65 199.199.125.28 L=119 S=0xC0 I=63767 F=0x0000 T=64 Jun 3 17:02:47 eth0-core0 kernel: IP acct in eth0 UDP 165.113.1.73:53 209.115.17.66:53 L=56 S=0x00 I=25895 F=0x0000 T=57 Jun 3 17:02:47 eth0-core0 kernel: IP acct out eth2 UDP 165.113.1.73:53 209.115.17.66:53 L=56 S=0x00 I=25895 F=0x0000 T=56 Jun 3 17:02:47 eth0-core0 kernel: IP acct out eth0 ICMP/3 209.115.17.65 165.113.1.73 L=118 S=0xC0 I=63769 F=0x0000 T=64 Jun 3 17:02:48 eth0-core0 kernel: IP acct in eth0 UDP 166.93.1.3:63098 209.115.17.66:53 L=56 S=0x00 I=44767 F=0x0040 T=245 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth2 UDP 166.93.1.3:63098 209.115.17.66:53 L=56 S=0x00 I=44767 F=0x0040 T=244 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth0 ICMP/3 209.115.17.65 166.93.1.3 L=118 S=0xC0 I=63770 F=0x0000 T=64 Jun 3 17:02:48 eth0-core0 kernel: IP acct in eth0 UDP 198.81.19.238:4569 209.115.17.66:53 L=59 S=0x00 I=34977 F=0x0000 T=20 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth2 UDP 198.81.19.238:4569 209.115.17.66:53 L=59 S=0x00 I=34977 F=0x0000 T=19 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth0 ICMP/3 209.115.17.65 198.81.19.238 L=121 S=0xC0 I=63771 F=0x0000 T=64 Jun 3 17:02:48 eth0-core0 kernel: IP acct in eth0 UDP 128.112.129.15:56224 209.115.17.66:53 L=58 S=0x00 I=50842 F=0x0040 T=247 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth2 UDP 128.112.129.15:56224 209.115.17.66:53 L=58 S=0x00 I=50842 F=0x0040 T=246 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth0 ICMP/3 209.115.17.65 128.112.129.15 L=120 S=0xC0 I=63772 F=0x0000 T=64 Jun 3 17:02:48 eth0-core0 kernel: IP acct in eth0 UDP 158.152.1.81:53 209.115.17.67:53 L=56 S=0x00 I=21310 F=0x0000 T=53 Jun 3 17:02:48 eth0-core0 kernel: IP acct out eth2 UDP 158.152.1.81:53 209.115.17.67:53 L=56 S=0x00 I=21310 F=0x0000 T=52 The thing that makes it "interesting" is the fact that most implementations DO send an ICMP unreach back. The ICMP Unreach traffic alone generated in the neighborhood of 1.7Mb before they routed the netblock in question to a loopback interface on the 7507. The attacker was sending less that 300Kb of traffic and consuming 2Mb. ------- John Fraizer (root) | __ _ | The System Administrator | / / (_)__ __ ____ __ | The choice mailto:root@EnterZone.Net | / /__/ / _ \/ // /\ \/ / | of a GNU http://www.EnterZone.Net/ | /____/_/_//_/\_,_/ /_/\_\ | Generation A 486 is a terrible thing to waste...

1 0

RE: Sprint's New ION products
by Pete Farmer 04 Jun '98

04 Jun '98

Nathan Stratton [mailto:nathan@robotics.net] wrote: >Right now most of the move to voice over packet/cell networks is >because people want to avoid termination and USF charges. The pace will >accelerate even more now that port density and port cost is approaching >that of TDM. For Sprint simply to use ATM in their core does not let them get around termination and USF. Only if they use an access mechanism other than standard PSTN do they get around these charges. Unfortunately, it's rather difficult to discern what the access mechanism will be for their 'ION' psudo-offering. Pete ______________________________________________________________________ Peter J. Farmer E-mail: pfarmer(a)strategies-u.com Strategies Unlimited Voice: +1 650 941 3438 201 San Antonio Circle, Suite 205 Fax: +1 650 941 5120 Mountain View, CA 94040 WWW: http://www.strategies-u.com

3 2

Re: PPP over Ethernet?
by woody＠zocalo.net 04 Jun '98

04 Jun '98

The standard for PPP over Ethernet is L2TP. The pre-standard proprietary implementations were L2F and PPTP, from Cisco and Microsoft, respectively. I believe you'll find Internet Drafts on all three. The idea is neither "highly bogus" nor used in association with 10Base-T cross-over cables. It's used to tunnel ATCP, IPCP, IPXCP, et cetera over IP. -Bill Woodcock

2 1