- Test - lists.nanog.org

NANOG 22 Hotel Information
by Carol Wadsworth 23 Mar '01

23 Mar '01

Meeting registration for NANOG 22 (May 20-22) will open on Monday, April 2. For those of you who want to make your hotel reservation in the interim, details follow: Meeting location and overnight accommodations: DoubleTree Paradise Valley Resort 5401 North Scottsdale Road Scottsdale, Arizona 85250 Room block name: NANOG Dates: May 19-22, 2001 Rate: $119 In-house Reservations: 480-947-5400 Central Reservations (tollfree): 800-222-TREE Note: The special rate will also apply if you want to stay any extra days prior to or following the meeting. The Call for Presentations announcement can be found at: http://www.nanog.org/mtg-0105/call22.html http://www.nanog.org/mtg-0105/call22.pdf The proposal deadline is April 9. Hope to see you in Scottsdale! Carol Wadsworth Merit Network, Inc.

1 0

Advanced Warning: Qwest/Lucent ATM upgrade
by Sean Donelan 23 Mar '01

23 Mar '01

Routing may be weirder than normal this evening. Qwest is upgrading their Lucent ATM switches to fix a bug, and needs to shutdown their entire ATM backbone for a couple of hours. Qwest hasn't responded to my e-mails asking what is happening, but other ISPs have outlined alternative routes they will be using. Most will try to maintain service by additional peering, deaggregated routes, or temporary third-party transit. Other providers use Lucent ATM switches. I don't know if they will also need to upgrade their networks at some time in the future to fix the same problem.

2 1

Real world Anti-DDOS attack practice.
by Yu Ning 23 Mar '01

23 Mar '01

Hi nanog, I'm sorry if I raise a cliché topic, but I've searched the nanog archive and get no satisfied answer. The question is quite simple, what's the best practice if my downstream customer report a heavy DDOS attack (icmp based, not source addr. spoofing)? Yes, to filter out the packet via ACL, but the impact on oc3/48 link with ACL packet filtering sounds to be a nightmare. If there is any effective practice to prevent my engineer from patching the router here and there via packet ACL ? Yes again via dCAR to rate-limiting the icmp traffic, but as soon as you mention the packet-filtering method, it seems as if my router is in smoke. Then I wonder what my US counterpart do to beat DDOS attack to their customer? Best real world practice ? How about tier-1 like UUNet ? thanks for any input. -------------------------------------------- (Mr.) Yu Ning, Chief Engineer ChinaNET Sr. Support & New Service Dev. Data Communication Bureau, China Telecom Beijing, P.R.China +86-10-62072357/62072354 --------------------------------------------

5 5

Re: Clear Channel on a T1
by Ron Buchalski 23 Mar '01

23 Mar '01

Good point, Phil. I believe that the modern T1 spec states that a maximum or 45 zeros can be sent before the CSU will inject a '1'. I don't have a spec handy to verify it. Of course, with B8ZS, you'll never approach this long string of zeros on the line, so it's a moot point. B8ZS replaces eight zeros with 4/4 ones/zeros, in a known bipolar violation pattern. This will insure that ones' density is met on the line, and the circuit endpoints can replace the injected ones with eight zeros and restore the bit patterns of the modified DS0s. It's also the source of grief when a telco provisions one segment of a T1 circuit without B8ZS. When this non-B8ZS segment sees too many zeros, ones will be injected in the stream, corrupting the data. Of course, this NEVER happens anymore...right? ;-) -rb >From: Phillip Vandry <vandry(a)TZoNE.ORG> >To: "Ron Buchalski" <rbuchals(a)hotmail.com> >CC: nanog(a)merit.edu >Subject: Re: Clear Channel on a T1 >Date: Thu, 22 Mar 2001 18:04:46 -0500 (EST) > > >Please correct me if I'm wrong (because I'm not _that_ certain of this) but >I think you still need the 1's density to maintain timing even in the >absence of repeaters. If you have a whole bunch of zero bits in a row the >voltage on the line will remain 0 for the length of the series. You get no >indication of when one bit ends and another starts. If the string of zeroes >is very long, the receiver might miscount the number of 0 bits that the >sequence represents (more than intended if its clock is running faster than >the sender's, less than intended otherwise). You don't have this problem >with a long string of 1 bits because the voltage transitioning from + to - >or back at every bit boundary. _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com

1 0

Re: rumors about rolling blackouts in East?
by Steven M. Bellovin 23 Mar '01

23 Mar '01

In message <Pine.SGI.4.10.10103231018451.2506393-100000(a)guinness.cs.stevens-tec h.edu>, Subhendu Ghosh writes: > > >I just heard a couple of folks talking about possible power problems in >the East this summer? - NY/NJ areas.. > >Any confirmation or just rumors.. There is some concern. New York City is planning on deploying ~10 "temporary" generators on barges around the city to provide extra generating capacity. --Steve Bellovin, http://www.research.att.com/~smb

1 0

rumors about rolling blackouts in East?
by Subhendu Ghosh 23 Mar '01

23 Mar '01

I just heard a couple of folks talking about possible power problems in the East this summer? - NY/NJ areas.. Any confirmation or just rumors.. -Subhendu

1 0

Re: Clear Channel on a T1
by CARL.P.HIRSCH＠sargentlundy.com 23 Mar '01

23 Mar '01

Here's a few more details - We provision T1. Salesdroid asks "You wanna give us $$$ for clear channel?" Us: "..... ?" Salesdroid: "It's so you can get 64kbit rather than the normal 56k" Us: "We thought all DS-0s were 64kbit?" Me: "I know! I'll ask NANOG!" Here's a document that further obfuscates the matter for me - http://www.comtest.com/tutorials/t1.html Anybody care to comment on its relevance in this situation? If I read it correctly, Clear Channel = Less Capacity. Doesn't sound right. I believe the carrier in question is AT&T. thanks for any input you guys might have -carl owner-nanog-outgoing @merit.edu AT nxmime To: CARL P HIRSCH/Sargentlundy, nanog(a)merit.edu AT nxmime@SNL-ccmail 03/22/01 12:12 PM cc: Subject: Clear Channel on a T1 Can anybody give me some info on what exactly "Clear Channel" means in relation to a DS1 circut? My gut tells me that it's only relevant with voice Ts... thanks, -carl hirsch network analyst sargent & lundy llc

6 5

Re: RADB lookup help
by John Fraizer 23 Mar '01

23 Mar '01

Thanks Saverio. Now for the next question. I'm working on the templates for all of our objects. Is there any preference as to which database people register their objects? --- John Fraizer EnterZone, Inc On Fri, 23 Mar 2001, Saverio Pangoli wrote: > > > On Fri, 23 Mar 2001, John Fraizer wrote: > > > I'm trying to do the equiv of "whois -h rr.arin.net -i origin as13944" at > > whois.radb.net but am not having any luck. Is there a way to form this > > query at whois.radb.net? > > > > I'm simply looking for all objects registered with an origin of as13944. > > Hi, > > If you don't mind querying the database manually, you can use the !g > command to get all the routes originating in as1394. > > Something like this: > > [saverio@stats saverio]$ telnet whois.radb.net whois > Trying 198.108.0.18... > Connected to whois.radb.net. > Escape character is '^]'. > !gas13944 > A226 > 209.41.244.0/24 216.29.112.0/20 216.226.152.0/24 216.29.116.0/24 > 216.29.120.0/2 > 1 209.41.207.0/24 209.41.247.0/24 209.115.38.0/24 209.115.39.0/24 > 209.115.53.0/2 > 4 209.115.120.0/22 216.28.0.0/21 216.28.151.0/24 216.226.128.0/19 > C > > I hope this helps... > > Cheers, > Saverio Pangoli (DANTE) >

1 0

Advice on colo locations in SJC
by Simon Lockhart 23 Mar '01

23 Mar '01

All, I'm just trying to work out a plan for building my network into the SJC area. (I've got LA sorted - it looks like 1 Wilshire is the place to be - or nearby!). I'll be needing to hook up to MAE-West ATM, and to PAIX, bringing in long-haul backbone circuits (from LA and probably an alternate route from NY), and hopping out a few local circuits to the Bay area. Initially I'd be looking for just 1 or two racks to house routers, but longer term would need access to a larger colo space if we build a server farm. All colo should be "neutral" where possible. MAE-W looks to be in 55 South Market, SJ, and a few people seem to have colo available there. Is that a good place to be? Any better suggestions? 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

RADB lookup help
by John Fraizer 23 Mar '01

23 Mar '01

I'm trying to do the equiv of "whois -h rr.arin.net -i origin as13944" at whois.radb.net but am not having any luck. Is there a way to form this query at whois.radb.net? I'm simply looking for all objects registered with an origin of as13944. Thanks, --- John Fraizer EnterZone, Inc

1 0