September 1998 - Test - lists.nanog.org

The Cidr Report
by Tony Bates 18 Sep '98

18 Sep '98

This is an auto-generated mail on Fri Sep 18 12:00:00 PDT 1998 It is not checked before it leaves my workstation. However, hopefully you will find this report interesting and will take the time to look through this to see if you can improve the amount of aggregation you perform. The report is split into sections: 0) General Status List the route table history for the last week, list any possibly bogus routes seen and give some status on ASes. 1) Gains by aggregating at the origin AS level This lists the "Top 30" players who if they decided to aggregate their announced classful prefixes at the origin AS level could make a significant difference in the reduction of the current size of the Internet routing table. This calculation does not take into account the inclusion of holes when forming an aggregate so it is possible even larger reduction should be possible. 2) Weekly Delta A summary of the last weeks changes in terms of withdrawn and added routes. Please note that this is only a snapshot but does give some indication of ASes participating in CIDR. Clearly, it is generally a good thing to see a large amont of withdrawls. 3) Interesting aggregates Interesting here means not an aggregate made as a set of classful routes. Thanks to xara.net for giving me access to their routing tables once a day. Please send any comments about this report directly to me. Check http://www.employees.org/~tbates/cidr-report.html for a daily update of this report. ------------------------------------------------------------------------------ CIDR REPORT for 18Sep98 0) General Status Table History ------------- Date Prefixes 110998 53238 120998 51812 130998 51865 140998 51883 150998 52067 160998 52072 170998 51947 180998 52123 Check http://www.employees.org/~tbates/cidr.plot.html for a plot of the table history. Possible Bogus Routes --------------------- AS Summary ---------- Number of ASes in routing system: 4020 Number of ASes announcing only one prefix: 1940 (1002 cidr, 938 classful) Largest number of cidr routes: 340 announced by AS3561 Largest number of classful routes: 1167 announced by AS701 1) Gains by aggregating at the origin AS level --- 18Sep98 --- ASnum NetsNow NetsCIDR NetGain % Gain Description AS701 1167 936 231 19.8% Alternet AS271 345 149 196 56.8% BCnet Backbone AS4293 442 255 187 42.3% IMCI AS3749 226 68 158 69.9% TECNET AS2493 367 212 155 42.2% iSTAR Internet, Inc. AS174 701 570 131 18.7% Performance Systems International AS4200 187 96 91 48.7% AGIS (Apex Global Information Ser AS5668 128 41 87 68.0% Century Telephone Inc. AS4755 124 46 78 62.9% Videsh Sanchar Nigam Ltd. India AS2685 227 150 77 33.9% IBM Global Network - US AS3804 231 158 73 31.6% Bell Solutions AS3221 117 44 73 62.4% EENet Autonomous System AS7046 246 175 71 28.9% UUNET-CUSTOMER AS4740 385 331 54 14.0% Ozemail Pty Ltd (ASN-OZEMAIL) AS549 203 150 53 26.1% ONet Backbone AS8517 124 72 52 41.9% ULAKNET-ASN AS10928 100 48 52 52.0% UNKNOWN AS6335 69 20 49 71.0% NTRNET AS72 87 39 48 55.2% Schlumberger Information Network AS1239 542 495 47 8.7% SprintLink Backbone AS803 78 34 44 56.4% SaskNet Backbone AS4539 68 28 40 58.8% NETROPOLIS AS852 160 121 39 24.4% AGT Advance Communication AS7122 119 80 39 32.8% MTS ADVANCED Network Policy AS6181 65 28 37 56.9% FUSE-NET AS10459 45 10 35 77.8% WANSASN AS2707 64 31 33 51.6% WEC AS1221 281 249 32 11.4% AARNET-AS AS6332 89 58 31 34.8% TELNOR AS4307 32 1 31 96.9% SVINET-1 For the rest of the previous weeks gain information please see http://www.employees.org:80/~tbates/cidr-report.html 2) Weekly Delta Please see http://www.employees.org:80/~tbates/cidr-report.html for this part of the report 3) Interesting aggregates Please see http://www.employees.org:80/~tbates/cidr-report.html for this part of the report

1 0

RE: How to loadshare over many E1 links
by Lutz, Kyle 18 Sep '98

18 Sep '98

Take a look at Kentrox's AAC-3 Access Concentrators and Multiplexers, I have used it in many T1-IMA solutions. http://www.kentrox/products/index.html -----Original Message----- From: Jesper Skriver [mailto:jesper@skriver.dk] Sent: Friday, September 18, 1998 2:57 AM To: nanog(a)merit.edu Cc: tdk-backbone(a)t.dk Subject: How to loadshare over many E1 links 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.

1 0

RE: hearing word of a fiber cut in virgina
by Goldstein_William＠bns.att.com 18 Sep '98

18 Sep '98

There IS a difference. The FCC keeps records of outages which must be reported by the carriers. It is worth checking these out. Bill Goldstein Senior Internet Specialist AT&T wgoldstein(a)att.com TEL:(412)642-7288 ---------- From: SEAN Sent: Thursday, September 17, 1998 2:44 AM To: nanog Cc: SEAN Subject: RE: hearing word of a fiber cut in virgina CMartin(a)mercury.balink.COM (Martin, Christian) writes: >Is it me, or has nearly every cut in the past 6 months been Worldcom >glass? Do they have any redundant rings? Looks like their growth has >been too agressive for sound SONET engineering efforts to be employed. >But, hey, they have a lot of glass... I don't have my papers at home, so this is my recollection. Over the last four years cable cuts and other physical outages seem pretty randomly distributed between AT&T, MCI, Sprint and Worldcom. Some years one carrier has more, other years a different carrier has more. However the affect of an outage has not been equally distributed. Each carrier has its strengths in different markets. Worldcom happens to be fairly strong in the Internet circuit market so you read about Worldcom cable cuts on Internet lists. If you were a financial type, you would have thought MCI had a lot of problems. Remember the cable cut in Manhatten earlier this summer, it affected some Internet traffic but the brokers on wall street were livid. If you were an airline, you would have thought AT&T had a lot of problems. Sprint has problems, but I don't know which market segment it is really strong in. Salespeople for each of the above companies are usually more than willing to give you information about the reliability problems of their competitors. Getting information about their own company is harder. I doubt backhoes or gophers will show any favoritism for or against cable installed new competitors such as Qwest or Level3. Never lose respect for Murphy's power. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation <<File: RE_ hearing word of a fiber cut in virgina.TXT>>

2 1

And we thought the text part of the Starr Report would be bad
by Christian Nielsen 18 Sep '98

18 Sep '98

http://dailynews.yahoo.com/headlines/wr/story.html?s=v/nm/19980918/wr/video… SAN FRANCISCO (Wired) - If Congress decides to release the videotape of President Clinton's grand jury testimony, the heavy load of video traffic could be the biggest test yet for the Internet's infrastructure. Christian

3 2

RE: How to loadshare over many E1 links
by Martin, Christian 18 Sep '98

18 Sep '98

I didn't know there was any full-duplex 10BaseT spec... -Chris > -----Original Message----- > From: Scott Whyte [mailto:swhyte@cisco.com] > Sent: Friday, September 18, 1998 11:40 AM > To: Jesper Skriver; nanog(a)merit.edu > Cc: tdk-backbone(a)t.dk > Subject: Re: How to loadshare over many E1 links > > > Jesper, > > You might consider MLP (multi-link PPP), with VIP2/50's. > > If you do go with an IMUX, I definitely recommend a HSSI to > the routers, rather than Ethernet, unless the Ethernet is > full-duplex. I've seen really poor performance due to > collisions on the Ethernets at either end because the middle > is actually 16Mbps full-duplex... > > Scott > > > At 12:57 AM 9/18/98 , Jesper Skriver wrote: > >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. > > > ---- > Scott Whyte |"Computers are different > from telephones. > Network Supported Accounts | Telephones ring. > Computers do not ring." > CCIE | -- > Andrew S. Tanenbaum >

2 1

RE: How to loadshare over many E1 links
by Martin, Christian 18 Sep '98

18 Sep '98

We're using the Stratacom MGX8220 - We can aggregate more than 8 with multiple cards... > -----Original Message----- > From: Jared Mauch [mailto:jared@puck.nether.net] > Sent: Friday, September 18, 1998 11:41 AM > To: Jesper Skriver; nanog(a)merit.edu > Cc: tdk-backbone(a)t.dk > Subject: Re: How to loadshare over many E1 links > > > This is a problem because igp issues: > > nnn-7202(config)#router ospf 1 > nnn-7202(config-router)#maximum-paths ? > <1-6> Number of paths > > Because of this, you are unable to hanle those multiple paths > past 6. > > You also have the same isues with isis: > nnn-7202(config)#router isis home1 > nnn-7202(config-router)#maximum-paths ? > <1-6> Number of paths > > I don't have a pdf viewer (on this machine), so can't cite this > right now, but I think that the dl imux will handle up to 8 E1's: > http://www.dl.com/online/datashts/dl3800e.pdf - jared On Fri, Sep 18, 1998 at 09:57:08AM +0200, Jesper Skriver wrote: > 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. -- Jared Mauch | pgp key available via finger from jared(a)puck.nether.net | http://puck.nether.net/~jared/

1 0

RE: How to loadshare over many E1 links
by Martin, Christian 18 Sep '98

18 Sep '98

Try making the routes across the links doubly-recursive. I don't know how well this works, but it is worth a try. For example: RTRA-Loopback0 IP 10.1.0.1/30 RTRA-Loopback1 IP 10.1.1.1/30 RTRA-Loopback11 IP 10.1.11.1/30 RTRB-Loopback1 IP 10.2.0.2/30 RTRB-Loopback1 IP 10.2.1.2/30 RTRB-Loopback1 IP 10.2.11.2/30 Then route to one loopback over 4-6 E1's and the other over the other 4-6 E1s Then, route the Loopback 11s over the two loopbacks. Again, I don't know if this would work, but it just popped into my head. Example Config RTRA int loopback 0 ip address 10.1.0.1 255.255.255.252 ! int loopback 1 ip address 10.1.1.1 255.255.255.252 ! int loopback 11 ip address 10.1.11.1 255.255.255.252 ! int s0 ip address 192.168.0.1 255.255.255.252 ! int s1 ip address 192.168.1.1 255.255.255.252 ! <snip> int s11 ip address 192.168.11.1 255.255.255.252 <snip> ! The first 6 E1s ip route 10.2.0.0 255.255.255.252 192.168.0.2 ip route 10.2.0.0 255.255.255.252 192.168.1.2 ip route 10.2.0.0 255.255.255.252 192.168.2.2 ip route 10.2.0.0 255.255.255.252 192.168.3.2 ip route 10.2.0.0 255.255.255.252 192.168.4.2 ip route 10.2.0.0 255.255.255.252 192.168.5.2 ! The other E1s ip route 10.2.1.0 255.255.255.252 192.168.10.2 ip route 10.2.1.0 255.255.255.252 192.168.11.2 ip route 10.2.1.0 255.255.255.252 192.168.12.2 ip route 10.2.1.0 255.255.255.252 192.168.13.2 ip route 10.2.1.0 255.255.255.252 192.168.14.2 ip route 10.2.1.0 255.255.255.252 192.168.15.2 ! ip route 10.2.11.0 255.255.255.252 10.2.0.2 ip route 10.2.11.0 255.255.255.252 10.2.1.2 ! ip route 172.16.16.0 255.240.0.0 null 0 ! router bgp 65555 neighbor 10.2.11.2 remote-as 65554 network 172.16.16.0 mask 255.240.0.0 Of course, you may have to process-switch the packets, or use CEF, which you mentioned. -Chris > -----Original Message----- > From: Jesper Skriver [mailto:jesper@skriver.dk] > Sent: Friday, September 18, 1998 3:57 AM > To: nanog(a)merit.edu > Cc: tdk-backbone(a)t.dk > Subject: How to loadshare over many E1 links > > > 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. >

1 0

Re: hearing word of a fiber cut in virgina
by Sean Donelan 18 Sep '98

18 Sep '98

Not too shabby. Fiber cut at 2:02pm CDT (3:02pm EDT, 12:02pm PDT), first NANOG posting within 21 minutes. You beat the first carrier notification by about 5 minutes. >From: steve(a)altrina.exodus.NET >Subject: hearing word of a fiber cut in virgina >Date: 16 Sep 1998 12:23:34 -0700 > >Hello, > > Seems to be a MFS fiber cut in Bowling Green, Virgina. This is all >the information I have at this time. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

6 6

what's channellized E1?
by Yu Ning 18 Sep '98

18 Sep '98

Hi friends, A very very basic question: What does "channellized E1" mean? What's its difference from a "normal" E1 connection? reply to my mail box, if you think such message is too basic posting on nanog, thanks. regards, Yu Ning -- ___________________________________________* Yu Ning ATM R&D Centre of BUPT (Beijing U. of Posts&Telecom) Beijing (ZIP:100876, MBox:147#), P.R.China Ideas ONLY reflect my own views,:-) ___________________________________________

2 2

While perusing BGP accountments today
by alex＠nac.net 17 Sep '98

17 Sep '98

*> 2.52.228.144/32 209.83.160.97 0 6347 1239 i Wee? not to be superstitious, but doesn't sprint own the mirror image of the about (144.228.x.x)? 1 nac-chq1-e0-0 (209.123.20.251) 6.556 ms 4.507 ms 4.913 ms 2 nac-nyc6-s3-2 (209.123.11.222) 208.609 ms 46.004 ms 52.967 ms 3 nac-brd1-h0 (209.123.11.190) 21.686 ms 7.134 ms 8.227 ms 4 GA0F8B.GR-1.NycmNY.savvis.net (209.83.160.97) 38.873 ms 21.764 ms 28.229 ms 5 sl-gw12-pen-10-0-T3.sprintlink.net (144.228.165.13) 19.487 ms 69.539 ms 63.810 ms 6 sl-bb10-pen-2-3.sprintlink.net (144.232.5.85) 19.669 ms 14.627 ms 20.308 ms 7 sl-bb7-pen-0-0-0.sprintlink.net (144.232.5.54) 19.116 ms 28.941 ms 54.179 ms 8 sl-bb5-chi-1-0-0.sprintlink.net (144.228.10.38) 93.912 ms 43.670 ms 54.402 ms 9 sl-bb12-chi-0-2.sprintlink.net (144.232.0.169) 41.690 ms 30.253 ms 108.610 ms 10 sl-bb2-chi-4-0-0.sprintlink.net (144.232.0.142) 60.862 ms 52.579 ms 38.913 ms 11 144.228.252.11 (144.228.252.11) 45.391 ms 50.220 ms 85.955 ms 12 sl-paterno-2-0-128k.sprintlink.net (144.228.52.2) 167.325 ms 55.033 ms 76.683 ms 13 sl-gw11-chi-9-1-TS5-128k.sprintlink.net (144.228.52.1) 70.834 ms 123.828 ms 77.876 ms 14 sl-paterno-2-0-128k.sprintlink.net (144.228.52.2) 58.633 ms 57.476 ms 61.027 ms 15 sl-gw11-chi-9-1-TS5-128k.sprintlink.net (144.228.52.1) 59.639 ms 53.384 ms 49.821 ms -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- ISPF, The Forum for ISPs by ISPs. October 26-28, 1998, Atlanta, GA. Three days of clues, news, and views from the industry's best and brightest. http://www.ispf.com/ for information and registration. Atheism is a non-prophet organization. I route, therefore I am. Alex Rubenstein, alex(a)nac.net, KC2BUO, ISP/C Charter Member Father of the Network and Head Bottle-Washer Net Access Corporation, 9 Mt. Pleasant Tpk., Denville, NJ 07834 Don't choose a spineless ISP; we have more backbone! http://www.nac.net -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

3 2