Weekly Routing Table Report

Mailman List Mirror (Read Only)
1

This is an automated weekly mailing describing the state of the Internet
Routing Table as seen from APNIC's router in Japan.
Daily listings are sent to bgp-stats@lists.apnic.net

For historical data, please see http://thyme.apnic.net.

If you have any comments please contact Philip Smith <pfs@cisco.com>.

Routing Table Report 04:00 +10GMT Sat 23 Feb, 2008

Report Website: http://thyme.apnic.net
Detailed Analysis: http://thyme.apnic.net/current/

Analysis Summary

2

[cc's drastically reduced]

3

David Conrad wrote:

[cc's drastically reduced]

Analysis Summary
----------------
BGP routing table entries examined: 246872
   Prefixes after maximum aggregation: 124953
   Deaggregation factor: 1.98

Yow. Last week it was 1.96 and this constitutes a significant increase. Anybody have an idea of what happened?

Cox has jumped about 2k in the past month for me. If I could ever figure out what the OIDs are for BGP peers I would graph it.

Justin

4

David Conrad wrote:

[cc's drastically reduced]

Analysis Summary
----------------
BGP routing table entries examined: 246872
   Prefixes after maximum aggregation: 124953
   Deaggregation factor: 1.98

Yow. Last week it was 1.96 and this constitutes a significant increase. Anybody have an idea of what happened?

Cox has jumped about 2k in the past month for me. If I could ever figure out what the OIDs are for BGP peers I would graph it.

Something is happening since the 21st of this month.

Check out the plot of BGP entries and ASes on the first page of http://www.cidr.report.org. Both the number of BGP advertisements and the number of ASes have increased noticeably in the past couple of days.

Geoff

Justin Shore wrote:

5

I'm guessing that should be

http://www.cidr-report.org/

6

Probably not the safest way to go about it. Might work out ok, but with the dramatic potential to blow something up, especially on a router with a full table. The bgp mib doesn't contain an object to tell you, summarized, how many prefixes a peer is sending. You would have to iterate the full routing table and pick out routes that originate from a specific AS. And then repeat it periodically to generate your time series.

You'd be better off parsing output from a zebra peer. Slightly hackish but far safer, and efficient.

- billn

7

Geoff Huston wrote:

----------------

BGP routing table entries examined: 246872
   Prefixes after maximum aggregation: 124953
   Deaggregation factor: 1.98

Yow. Last week it was 1.96 and this constitutes a significant increase. Anybody have an idea of what happened?

Cox has jumped about 2k in the past month for me. If I could ever figure out what the OIDs are for BGP peers I would graph it.

Something is happening since the 21st of this month.

Check out the plot of BGP entries and ASes on the first page of http://www.cidr.report.org. Both the number of BGP advertisements and the number of ASes have increased noticeably in the past couple of days.

We need ~2500 prefixes a week in order to hit 500,000 routes in the dfz by this time in 2010 (for those of you working on your graph curve fitting exercises). I don't see people's perceived or real TE-needs or address exhaustion pressure lightening up in the immediately foreseeable future. The question of which provider it is this week is interesting but not actually germain to overall table growth.

8

No, but the BGP4+ MIB does… different vendors implemented that MIB early with differing OIDs :frowning:
Support for the correct OID is, er, sadly lacking in common versions of popular routing OSen.