From: randy@psg.com (Randy Bush)
Date Prefixes 171196 40844 171196 40868 181196 40933 191196 42076 201196 40674 211196 40583 221196 40525
Tony, I suspect that you and your weekly report are, in a large part, responsible for this. I thank you. Our routers thank you.
Date Prefixes 280996 42250 290996 42040 300996 42235 011096 42228 021096 42249 031096 42109 041096 41894 Yeah, me, too! Comparing with nearly 2 months ago pasted in above, this is a fantastic trend! Down 1700 routes (except sometimes).
Frank, if you're out there, has the curve changed since Tony started publishing again?
One of the things that I'm curious about: have any of the "interesting aggregates" been acted upon? The list seems to be getting _more_ entries, instead of the fewer I'd expect. Perhaps I don't understand the algorithm. I find it hard to compare and check by eye. By simple diff, I can see that ANS and UUNET have been fixing them, since they've mostly disappeared from the list. But others seem to pop in and out, and sometimes spawn many where there was only one line before, particularly AGIS, MSEN, PREPnet and PSInet, all of whom should know better. Could that be automatically added to the weekly summary? "Recent aggregations", just before Interesting? A tally of successes that would be more clear than withdrawn route counts.... But folks _are_ withdrawing routes by the hundreds! WSimpson@UMich.edu Key fingerprint = 17 40 5E 67 15 6F 31 26 DD 0D B9 9B 6A 15 2C 32 BSimpson@MorningStar.com Key fingerprint = 2E 07 23 03 C5 62 70 D3 59 B1 4F 5E 1D C2 C1 A2
"William Allen Simpson" <wsimpson@greendragon.com> writes: * > From: randy@psg.com (Randy Bush) * > * > > Date Prefixes * > > 171196 40844 * > > 171196 40868 * > > 181196 40933 * > > 191196 42076 * > > 201196 40674 * > > 211196 40583 * > > 221196 40525 * > * > Tony, I suspect that you and your weekly report are, in a large part, * > responsible for this. I thank you. Our routers thank you. * > * Date Prefixes * 280996 42250 * 290996 42040 * 300996 42235 * 011096 42228 * 021096 42249 * 031096 42109 * 041096 41894 * * Yeah, me, too! Comparing with nearly 2 months ago pasted in above, this * is a fantastic trend! Down 1700 routes (except sometimes). * Well - let's not get too excited. Check the last few days out. Date Prefixes 201196 40674 211196 40583 221196 40525 231196 40687 241196 40908 251196 40844 261196 41043 Soon creeps back up although some folks certainly have been cleaning up a fair bit recently. EMBRATEL and BBN Planet South Eastern Region are two that I noticed. I've added a little extra piece to the report which is a plot tracking the history since I've restarted this project. Check http://www.employees.org/~tbates/cidr.plot.xbm This will be updated on a daily basis. This will be referenced on the weekly mail update as well as the main web page http://www.employees.org/~tbates/cidr-report.html * * > Frank, if you're out there, has the curve changed since Tony started * > publishing again? * > I'd be interested in this also ;-). * One of the things that I'm curious about: have any of the "interesting * aggregates" been acted upon? The list seems to be getting _more_ * entries, instead of the fewer I'd expect. Perhaps I don't understand * the algorithm. * The algorithm just looks for prefixen which are subnets of old classful routes using the assumption that today there is little reason to be announcing these and in most cases with the ones it finds this seems to be the case. I'm not sure either why this list keep growing becuase even just eye-balling them, most dont seem to make sense. The most common is /24s from a legacy Class B. It also has the benefit of often pointing out obvious leaks (like /32s from time to time) * I find it hard to compare and check by eye. By simple diff, I can see * that ANS and UUNET have been fixing them, since they've mostly * disappeared from the list. But others seem to pop in and out, and * sometimes spawn many where there was only one line before, particularly * AGIS, MSEN, PREPnet and PSInet, all of whom should know better. * As a start I've modified this to be sort by AS rather than prefix to perhaps help a little and added some description of the network name also (this was based on a suggestion from Mike Patton). This is not really how I'd like it but its a start. It now looks like this. List of possibly interesting aggregates --------------------------------------- Aggregate | Origin | AS Description | Netname ------------------------------------------------------------------------------- 154.11.1.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.2.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.8.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.16.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.24.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.25.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.32.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.36.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.37.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.48.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.64.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.96.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.97.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.98.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.99.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.108.0/22 | AS8013 | PSINET-CA | Performance Systems Int 154.11.108.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.109.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.110.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.126.0/24 | AS8013 | PSINET-CA | Performance Systems Int 154.11.127.0/24 | AS8013 | PSINET-CA | Performance Systems Int 155.229.30.0/24 | AS7330 | INTERNETLABS | Internet Atlanta, Inc. 155.229.176.0/21 | AS7045 | WEATHER | * 12.0.0.0/9 | AS7018 | AT&T worldnet network | AT&T ITS (NET-ATT) 168.234.128.0/24 | AS6458 | EMPRESSA GUATEMALTECA D | Universidad del Valle d 168.234.135.0/24 | AS6458 | EMPRESSA GUATEMALTECA D | Universidad del Valle d 168.234.136.0/22 | AS6458 | EMPRESSA GUATEMALTECA D | Universidad del Valle d 170.119.27.0/24 | AS6335 | NTRNET | LG&E Energy (NET-LGE-CO 170.119.229.0/24 | AS6335 | NTRNET | LG&E Energy (NET-LGE-CO 149.236.92.0/24 | AS6292 | FCI | Bruker Analytische Mess [...] * Could that be automatically added to the weekly summary? "Recent * aggregations", just before Interesting? A tally of successes that * would be more clear than withdrawn route counts.... * Okay let me think about this a while. --Tony
participants (2)
-
Tony Bates -
William Allen Simpson