- NANOG - lists.nanog.org

Re: Napster and others...
by smd＠clock.org 09 Mar '00

09 Mar '00

| What would be your suggestions on making Napster a more network | friendly app? 1. RFC 2001 compliance. Keep it TCP. 2. Although RFC 1349 is supposedly dead and the TOS octet in the RFC 791 scheme is dead too[*], it is at least good politics to set a low TOS value on the bulk transfer traffic. (If not on all traffic). Thus, routers configured to do TOS-based fancy queueing will DTRT and fewer people will accuse Napster of being a resource pig. 3. "If we aren't network friendly, please let us know what will make us more network friendly" is a great attitude to demonstrate. Hopefully this will be appreciated by actual operators (at least the ones who don't pay per-packet/per-byte charges). Sean. [*] RFC 2474/RFC 2475 but don't hold your breath -:) Who implements this could be a NANOG topic. -:) -:)

7 6

Provider-local addresses?
by Brian Candler 09 Mar '00

09 Mar '00

Hello, Having watched the 'route filtering panel' discussion from the most recent NANOG meeting, I have an idea which I would like to float. One of the concerns raised was that small ISPs are not eligible for provider-independent blocks; unfortunately this forces them to renumber if they change upstream provider. It was pointed out that this in itself is not a good reason for getting PI space, because then everybody would want it. Now, renumbering is tedious but in principle not difficult. The major difficulties are with DNS - in particular, changing the DNS settings in all the clients, which may well have the IP addresses of the ISP's caching nameservers hardcoded in. [1] My suggestion is that a small amount of unrouted space is allocated for "provider-local" use. For sake of example, let's say 200.0.0/24. Then we can assign say 200.0.0.1 and 200.0.0.2 as well-known addresses for local primary and secondary DNS caches. These are configured as loopback interfaces on the ISP's DNS servers, and static /32 routes injected into the provider's IGP to reach them. [2] The ISP would not need to divulge the 'real' addresses of its DNS caches, and could thus renumber at will. It would also benefit customers who roam between multiple ISPs, as they wouldn't have to keep changing their client settings. Notice it should work for multiple levels of service-provider hierarchy: sending a packet to 200.0.0.1 would follow defaultroute until it hit the first ISP who implements a server on this address. These addresses might be useful for other services too: e.g. outgoing SMTP relays, NTP servers, etc. In principle all these should have DNS names which customers use, but experience shows that customers _do_ end up hardcoding IP addresses into their configurations, which makes it very difficult to renumber them. The allocation would only need to be a single /24 or smaller, and it could come from any useless bit of 'toxic waste' because it would not actually be routed on the Internet. Does anybody see any value in this proposal? Of course, it's not much help to an ISP with downstream leased-line customers, because the customers will have to renumber as well. Regards, Brian Candler. [this posting is from my personal address and not on behalf of my employer] [1] There are MS extensions to PPP to negotiate primary/secondary DNS cache, but not everybody uses them. [2] We can also preallocate space in a way suitable for subnetting: 200.0.0.1 Primary DNS cache 200.0.0.2 Primary NTP server ... 200.0.0.129 Secondary DNS cache 200.0.0.130 Secondary NTP server ... This allows people to implement this proposal by bringing up interface aliases, without messing with host routes in IGP, and still have some resilience by putting the primary/secondary services on different physical networks.

1 0

Lost a URL to BindView
by Henry R. Linneweh 09 Mar '00

09 Mar '00

hmm bookmarks took a dump and are not recoverable I am missing the link to Bindview with that paper on Razor -- Thank you; |--------------------------------------------| | Thinking is a learned process so is UNIX | |--------------------------------------------| Henry R. Linneweh

1 0

question about MPOA
by Jane Chen 09 Mar '00

09 Mar '00

Is there any ATM switch or IP router wiht ATM interface which support MPOA (RFC1483) using VC multiplexing scheme? Is there any Service provider use VC multimexing in ATM interface to carry IP traffic? Thanks. Jane

1 0

ALTDB DNS problem.
by Steve Rubin 09 Mar '00

09 Mar '00

nice to see that quality control at Network Solutions has not improved. soapbox: {118} % whois altdb.net Whois Server Version 1.1 Domain names in the .com, .net, and .org domains can now be registered with many different competing registrars. Go to http://www.internic.net for detailed information. No match for "ALTDB.NET". I apologize to the users of ALTDB in advance, I am not sure when this disapeared, it was pointed out to me just a few minutes ago. ...And yes, I did pay the bill. -- Steve Rubin * ser(a)tch.org * http://www.tch.org/~ser/

1 0

NSI Bought Out
by Henry R. Linneweh 08 Mar '00

08 Mar '00

Network Solutions: It will be part of one-stop shop after purchase by VeriSign. http://www.mercurycenter.com/svtech/news/indepth/docs/veri030800.htm -- Thank you; |--------------------------------------------| | Thinking is a learned process so is UNIX | |--------------------------------------------| Henry R. Linneweh

1 0

DS3 multiplexors
by Simon Lockhart 08 Mar '00

08 Mar '00

Can anyone recommend a product which can multiplex onto a DS3 or E3 the following services: - Fractional DS3/E3 (G.703 and/or HSSI) - T1/E1 (G.703 and/or X.21) - E2 (G.703) (Optional) - 64k channels (X.21) (Optional) At the moment we're using Digital Link DL3100's, but are hitting problems with having just 3 HSSI ports. Thanks, Simon -- Simon Lockhart | Tel: +44 (0)1737 839676 Internet Engineering Manager | Fax: +44 (0)1737 839516 BBC Internet Services | Email: Simon.Lockhart(a)bbc.co.uk Kingswood Warren,Tadworth,Surrey,UK | URL: http://support.bbc.co.uk/

1 0

Re: AGIS Bites the dust
by Sean Donelan 07 Mar '00

07 Mar '00

>On Mon, 06 March 2000, "Henry R. Linneweh" wrote: >> AFTER ROUGH VOYAGE AGIS HITS THE ROCKS >> http://www.zdnet.com/intweek/stories/news/0,4164,2455629,00.html >From the article: >Founded in June 1994 by Philip Lawlor, AGIS is one of the six original >backbones tapped by the National Science Foundation to carry commercial >Internet traffic. To be pedantic, AGIS was never selected or tapped by the National Science Foundation to carry commercial Internet traffic. No provider was. Any and all providers who connected to the three primary NAPs (new york, chicago, and san francisco) were considered NSPs. Even the term "original" isn't true since there were other commercial networks who never met the definition of NSP, such as BBN, CERFNET and UUNET. Heck I think even Net99 was bigger than AGIS, before it self-destructed and Phil bought them.

2 1

Napster and others...
by Jeff Kell 07 Mar '00

07 Mar '00

Is there an updated list of ports used by Napster and various games (Doom, Unreal, etc)? At this point we're just collecting data points and not actively filtering anything, but of course the IANA port number list is sadly out of date with regard to these tools. Thanks in advance, Jeff Kell <jeff-kell(a)utc.edu> Systems/Network Administrator University of Tennessee at Chattanooga

10 13

Napster and others...
by Scott McGrath 07 Mar '00

07 Mar '00

1 0