On Fri, Sep 22, 2000 at 03:44:50PM -0400, Vijay Gill wrote:

On Fri, 22 Sep 2000, Kai Schlichting wrote:

[ http://www.employees.org/~tbates/cidr.plot.html for a plot ]

...

of engineering experience announcing 32 /24's out of their scattered POPs in addition to their /19 aggregate because they can't get their load balancing right without it or have never heard of the 'NO_EXPORT' attribute (I could point fingers now)?

I suspect data from inside providers with aggressive filtering policies might prove to be interesting.

Some promising local providers were past the 150k mark a while ago for sum (internal + global bgp) routes and were growing superlinearly the entire time because of the ever increasing number of customers.

I have in one location over 100k prefixes received from one nice large international provider: Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd a.b.c.d 4 701 10776831 107250 13024503 0 0 1w4d 105250 My primary concern is that the current growth is going to outscale the medium sized providers as the one vendor that the above output came from doesn't appear to be providing enough memory in their equipment, or have enough memory options available. I am not predicting "the internet will end in 9 days, repent", but it is some cause for concern in the next 6-12 months. What would be nice is if someone were to put together an auto-nagger similar to the "your dns server is lame" of the old days messages. Back when the network tables were about 40-45k, I would read the output of "sh ip b" and send messages to people asking them to aggregate, but those days are over... (at least for me..) - jared -- Jared Mauch | pgp key available via finger from jared@puck.nether.net clue++; | http://puck.nether.net/~jared/ My statements are only mine. END OF LINE | Manger of IP networks built within my own home

Don't forget my favorite target for 'aggregation agression', dear ol' UUNet, just for grins I polled one of my border routers, 1784 prefixes being heard with an AS path of ^701$, of which 271 are /24's, not to mention the rest of their unaggregated announcements. This apparentl lack of concern for how they affect the rest of the internet with their lasse fair attitude when it comes to making their customers announce properly to them (and thus to us) is simply amazing. ----- Original Message ----- From: "Kai Schlichting" <kai@pac-rim.net> To: <nanog@merit.edu> Sent: Friday, September 22, 2000 3:38 PM Subject: exponential route prefix growth, was: Re: The Cidr Report

At Friday 03:00 PM 9/22/00, Tony Bates wrote:

...

Check http://www.employees.org/~tbates/cidr.plot.html for a plot of the table history.

Is this just me, or has noone commented on this in a while? Somehow we are seeing an exponential growth in the number of prefixes in the last 12 months, while it was linear for the 5 years before that ( http://www.employees.org/~tbates/cidr.hist.plot.html )

What 'threshold' has triggered this sudden event, with routes going from 60,000 to 90,000 in just 12 months? Multihoming becoming fashionable? Dinky-rink providers getting multihomed, and for lack of engineering experience announcing 32 /24's out of their scattered POPs in addition to their /19 aggregate because they can't get their load balancing right without it or have never heard of the 'NO_EXPORT' attribute (I could point fingers now)?

At 15:38 09/22/2000 -0400, Kai Schlichting wrote:

At Friday 03:00 PM 9/22/00, Tony Bates wrote:

...

Check http://www.employees.org/~tbates/cidr.plot.html for a plot of the table history.

Is this just me, or has noone commented on this in a while? Somehow we are seeing an exponential growth in the number of prefixes in the last 12 months, while it was linear for the 5 years before that ( http://www.employees.org/~tbates/cidr.hist.plot.html )...

Actually, I have been looking at Tony's plots of active AS numbers for some time, and every time I do, I get a nervous twitch in my gut. Thanks for your post, Kai... You inspired me to do the obvious analysis. It turns out that the trend did not suddently become exponential; rather, it was exponential all the time, as near as I can figure. Tony was kind enough to ship me his source data. I cleaned up the obvious outliers (significant deviation from monotonic non-decreasing, i.e. the big dips in his graph) by hand and ran a quick exponential regression. The R-squared (goodness of fit) for the full sequence going back to 1996 comes out at .9957 - as good a fit as you are ever likely to see on real data. It's a beautiful, textbook result. Exponential! If I did not fat finger the math, the implications are obvious, and troubling. - Scott (speaking only for himself)

I'm tossing in my hat as a company who has recently multi-homed our enterprise network. We would have preferred to receive a continous block, /23 from ARIN. Unfortunatly they do not allocate smaller then a /20. We went to our provider and tried to justify a /23. They rejected our claim, even though our predicte growth puts us around 200 used IPs in a year. We ended up getting only one /24. Then another upstream provider, quite large, forced another /24 upon us. When we stated we didn't need/want it, they said they could take it back but it was not standard practice; all DS3 customers get a /24. Anyway... Think of all the other companies out there who get treated like this? Have you ever checked this URL out: http://www.employees.org/~tbates/checkas.html, select "Print Full Aggregation by AS Report." If you run this, almost 10,000 routes could be aggregated. This is a 11% savings! I've ran across content providers who have a /18, but announce them all as /24. That's 63 to many routes in my table. Minimum Savings: - UUNet CA (816): 137 - AT&T (7018): 67 - UUNet (701): 66 - Sprint CA (3602): 56 - Qwest (209): 44 - Genuity (1): 40 - Level 3 (3356): 17 Has there ever been consideration to create a group/organization to monitor these tables? Someone who can call these providers and enforce aggregation? Just a thought? Perhaps ICANN, ARIN, or someone could establish a team deticated to making sure the little guys don't get kicked out of the multi-homed world.

-----Original Message----- From: owner-nanog@merit.edu [mailto:owner-nanog@merit.edu]On Behalf Of William Waites Sent: Friday, September 22, 2000 4:48 PM To: kai@pac-rim.net Cc: nanog@merit.edu Subject: Re: exponential route prefix growth, was: Re: The Cidr Report

Kai Schlichting <kai@pac-rim.net> wrote:

...

What 'threshold' has triggered this sudden event, with routes going from 60,000 to 90,000 in just 12 months? Multihoming becoming fashionable? Dinky-rink providers getting multihomed, and for lack

Fashionable or not, multihoming is a usefull and sound practice. The problem is that regulatory organizations (ex. ARIN) make it very difficult to do it properly, and so cause inefficient use of address space and routing table bloat.

-w

up getting only one /24. Then another upstream provider, quite large, forced another /24 upon us. When we stated we didn't need/want it, they said they could take it back but it was not standard practice; all DS3 customers get a /24. Anyway... Think of all the other companies out there who get treated like this?

i would not make the argument that this is good, best practice, beneficial, ... but i have to whine that it is *very* hard in a large provider to make it clear to and easy for sales and provisioning folk to adequately determine what a customer can actually justify (in an rir sense) and provision thusly. to us it may be obvious that a residential dsl customer might have a bonkers gadget-crazed geek population that actually warrants a /25 while an oc3 turn-up who insists on a natting firewall only needs a /29. this is very hard to explain to a normal human being in sales/provisioning who has to do 20-50 of them a day, is underpaid, and who is completely disconnected from the physics and consequences of address allocation policy. randy