- Test - lists.nanog.org

CHI contact issue..
by Richard Irving 19 Sep '98

19 Sep '98

Thanks gentlemen, CHI contacted me immediately, (I love the nanog list as a back up contact :). According to on-site technicals there is a fiber cut at "Midwest Relay", I am not sure how many yet. DS3's... But, emergency circuits are getting priority at 1 hour 30 minutes est. Have a nice saturday. :(

1 0

Re: How to loadshare over many E1 links
by Hank Nussbacher 19 Sep '98

19 Sep '98

At 11:56 AM 9/18/98 -0400, Leo Bicknell wrote: On a slightly related matter, does anyone know how to load balance non-parallel links to a BGP peer? Thanks, Hank >In article <19980918095708.B10051(a)skriver.dk> you write: >>How would you loadshare over many (>6) parallel E1 links. Currently we >>do it by connecting them directly to the Cisco's in each end, and do >>CEF based per-packet loadsharing, it works fine, but support a max of >>6 E1's ... > > Cisco has a nice white paper on this: >http://www.cisco.com/warp/public/733/adap/multi/althb_wp.htm > > I think Multilink PPP may be your answer, as it goes >up to 8 links. There is a CPU hit though, I've never tried >Multilink on a high speed interface so I don't know how bad >it might affect the router. > >-- >Leo Bicknell - bicknell(a)dimension.net >Network Engineer (CCIE #3440) - Dimension Enterprises >1-703-709-7500, fax, 1-703-709-7699 > >

1 0

Re: How to loadshare over many E1 links
by shields＠crosslink.net 19 Sep '98

19 Sep '98

In article <c=US%a=_%p=BAIS%l=MERCURY-980918163103Z-10472(a)mercury.balink.com>, "Martin, Christian" <CMartin(a)mercury.balink.com> wrote: > I meant the latter. I knew that there were switch to switch FDX > capabilities - I'd never seen a router interface or NIC though. Cool. For PC platforms, I've been happy with both Tulips and Intel eepro100b's at fdx 10BT. Both cards are real cheap and have 100BT support as well. Catalyst 1900s do fdx 10BT, as do 3com 1000s (but you'll want current firmware). I don't think Cisco has fdx 10BT interfaces but the new-style (VIP-type) FEIP2 will do fdx 100BT. I'm not sure if hdx/fdx autonegotiation works at 10BT -- I don't trust it anyway. -- Shields, CrossLink.

2 1

RE: got any good documents ?
by Goldstein_William＠bns.att.com 19 Sep '98

19 Sep '98

Seem to remember looking over it vaguely. As I remember, it dealt more with LAN than WAN issues and had little on MPOA. Bill Goldstein Senior Internet Specialist AT&T wgoldstein(a)att.com TEL:(412)642-7288 ---------- From: pat Sent: Friday, September 18, 1998 10:12 PM To: Goldstein, William Cc: pat; jamie; kbrown; nanog Subject: RE: got any good documents ? Any comments on this one: Tcp/Ip over Atm : A No-Nonsense Internetworking Guide by Berry Kercheval ... particularly from the view of ATM-based enterprise backbone running IP. At 12:05 PM 9/16/98 , Goldstein_William(a)bns.att.com wrote: > Dan Minoli has published a couple of good books on ATM this year. > > Also, Darren Spohn and Dave McDysan have a new one coming out in a few > weeks that is supposed to be very comprehensive. -- Patrick Lee <pat(a)patlee.org> <<File: RE_ got any good documents _.TXT>>

2 1

Question- Contacts at CHI-AADS during weekend
by Richard Irving 19 Sep '98

19 Sep '98

Hrmmmmm, Seems like the CHI-AADS emergency contacts have changed... At least for the weekend... Anyone have new outage contacts, for a Saturday ???

2 1

Re: FCC outage reports
by shields＠crosslink.net 19 Sep '98

19 Sep '98

In article <980918150644.eb48(a)SDG.DRA.COM>, Sean Donelan <SEAN(a)SDG.DRA.COM> wrote: > FCC outage reports by year and inter-exchange carrier > > 1992 1993 1994 1995 1996 > AT&T 3 15 7 18 17 > MCI 13 12 19 9 15 > Sprint 3 11 3 12 9 > WorldCom NA NA NA 7 13 > > There are all sorts of normalization problems, different reporting [...] You'd also want to look at more than just the count of outage reports; the size of the outages matters, the total size of the network, and the MTTR. If you "had to" distill it to a single number I suppose percentage uptime is the one to use. I'm not sure fiber cuts are common enough to make a meaningful statement about the reliability of one carrier vs. another. -- Shields, CrossLink.

1 0

How to loadshare over many E1 links
by Jesper Skriver 19 Sep '98

19 Sep '98

Hi, How would you loadshare over many (>6) parallel E1 links. Currently we do it by connecting them directly to the Cisco's in each end, and do CEF based per-packet loadsharing, it works fine, but support a max of 6 E1's ... I've been thinking om something similar to the Larscom inverse MUX (http://www.larscom.com/t3ft3/t3_megae.htm) but this one only support 4 E1's, then I can use multiple i-mux's and loadshare over the 8M links they provide, but it seems like a poor solution. Yes, I do know that a E3 would be a far better choice, but our ADM (Add-Drop-Multiplexer??) at this specific location only support E1's :-( /Jesper -- Jesper Skriver (JS249-RIPE), Network manager Tele Danmark DataNet, IP section (AS3292) One Unix to rule them all, One Resolver to find them, One IP to bring them all and in the zone to bind them.

5 4

RE: got any good documents ?
by Goldstein_William＠bns.att.com 19 Sep '98

19 Sep '98

Dan Minoli has published a couple of good books on ATM this year. Also, Darren Spohn and Dave McDysan have a new one coming out in a few weeks that is supposed to be very comprehensive. Bill Goldstein Senior Internet Specialist AT&T wgoldstein(a)att.com TEL:(412)642-7288 ---------- From: jamie Sent: Wednesday, September 16, 1998 10:04 AM To: kbrown Cc: nanog; jamie Subject: Re: got any good documents ? "ATM for Dummies". Honestly, it's a really good book. Bought it a year or so? ago. It's pretty thorough in explaining concepts. Don't see it on my bookshelf here, so I don't have an ISBN for ya. :( -jamie kbrown(a)teamhuber.com wrote: > From errors-nohumans(a)merit.edu Tue Sep 15 14:52:52 1998 > From: kbrown(a)teamhuber.com > X-Lotus-FromDomain: TEAMHUBER > To: nanog(a)merit.edu > Message-ID: <86256680.006BE767.00(a)domino.teamhuber.com> > Date: Tue, 15 Sep 1998 14:38:34 -0500 > Subject: got any good documents ? > Content-Disposition: inline > Sender: owner-nanog(a)merit.edu > > > I am preparing to give a presentation to end-users about using ATM > technologies in their environments. I'm doing well, with the exception of > finding a way to "dumb-down" ATM addressing. Might any of you have already > dealt with this issue...say perhaps with your managers? :=) > > If you have a document which might assist me with reducing the overhead > required to explain this, and you don't mind me borrowing some of the > context... please zap it my way. It will be greatly appreciated. > > Kevin Brown > kevin(a)teamhuber.com > > -- jamie rishaw (efnet:gavroche) American Information Systems, Inc. Tel:312.425.7140, FAX:312.425.7240 Help stop spam! router(config)#no ip routing thirty thousand feet above the earth..youre a beautiful thing.. <<File: Re_ got any good documents _.TXT>>

2 1

Re: More on video
by Sean M. Doran 19 Sep '98

19 Sep '98

Sandra Gittlen wrote: | In addition to the sheer size of the file, one issue is how to | maintain a user's link for up to four hours. The Internet was | not designed for such stateful connections. Stateful connections? Do you mean "long-lasting bulk-transfers"? If so, then, while the current Internet is not optimal for them (in large part because of the sheer amount of very short flows that in aggregate do not do particularly good jobs of congestion avoidance, and because of open loop non-congestion-avoiding traffic), it certainly was designed to support them, and they can and do happen. Multi-week TCP connections (ssh/telnet/X) across the Internet are nothing very new, and they work just fine. Really large TCP bulk transfers lasting hours and even days is nothing new in some places (supercomputer centres and backup storage facilities come to mind). Multi-hour multimedia stuff is not very new either -- people watching IETF multicasts is not very new, and NANOG unleashed the concept of watching several hours of realvideo too. What is different here is that lots of people will be trying to watch several hours of streaming realvideo all at once, and there is a risk of something breaking down -- more likely the realvideo servers won't cope with the load. The really big pity is that TCP mode isn't the default; shoving non-real-time realvideo into a TCP stream solves the worries about realvideo's UDP-mode congestion avoidance not being _no more aggressive than the congestion avoidance detailed in RFC 2001_ due to implementation bugs or (intentional or accidental) protocol misdesign. If, however, realvideo does RFC-2001-style congestion avoidance properly, then the core of the network won't even notice Monday, since the long-lived congestion-avoiding flows are "nicer" to bottlenecks than the hordes of short-lived flows one normally sees when people are surfing the web. Sean.

1 0

Re: More on video
by brandon＠rd.bbc.co.uk 18 Sep '98

18 Sep '98

> Just did a story on this yesterday... > > http://www.nwfusion.com/news/0917starr.html > > 'Net braces for Clinton testimony > > By Sandra Gittlen > Network World Fusion, 9/17/98 > > In London, the BBC's online venture is also gearing up. BBC Online has > capacity for 20,000 concurrent videostreams, said marketing editor > Keith Roberts. Err, 10K license but that doesn't g'tee there's enough b/w for it. If it was a 28K stream that'd be 280M, we have around 145M total for this. If it was just a 5K speech service it'd be nicer. Speaking to a number of ISP & large users the other week there were conflicting opinions on this: ISP - Multicast doesn't scale, will cost too much to roll out, doesn't have an obvious pricing model and so on so we're not going to do it. Plus b/w will be so cheap it we won't care (assumes the Telcos pass on that low cost) Users - this is crazy, we should have multicast we can't keep burning so much b/w sending the same packets down the same wires > The BBC has some experience with this sort of massive > demand. Last August, the still nascent BBC Web effort was hard hit by > mourners seeking footage of Princess Diana. Although the site did not > crash, it did slow down, Roberts said. Yeah but I cheated and had 8 ISP's, RealNetwork & AOL help out by rebroadcasting our stream. > The BBC hosts its own RealPlayer video and audio servers separate from > the main site's Web servers. "That way, if there is enormous pressure > for video and audio, the Web servers for the site will not be > affected," he said. True the News web servers are on a seperate link but the Real servers share with other services... brandon

1 0