Re: [Sidrops] [GROW] Deprecation of AS_SET and AS_CONFED_SET -- feedback requested

[Warren Kumari <[e-mail]>]

Thank you Sriram and Lilia -- this is helpful / useful.

I wonder if the people who configured the 21 "Routes that seem
meaningful" still know that they are doing this, or if it is a lurking
horror....

W

On Wed, Oct 23, 2019 at 11:45 AM Sriram, Kotikalapudi (Fed)
<[e-mail]> wrote:
>
> Earlier in this thread, there was discussion about some measurements that were shared  (by Jared, Warren, myself).
> Jeff's suggestion: why don't we list all routes with AS_SET in addition to the stats.
> Now we have a detailed analysis (thanks once again to Lilia Hannachi my colleague at NIST):
> https://www.nist.gov/sites/default/files/documents/2019/10/23/detailed-as_set-analysis.txt
>
> The analysis includes the following:
> 1. Summary stats
> 2. List of ASes (AGGREGATORs) Ranked according to # Routes with AS_SET
> (including the organization name of the AGGREGATOR)
> 3. List of routes with AS_SET that seem meaningless or malformed.
> 4. List of routes with AS_SET that seem meaningful.
> 3. List of all routes with AS_SET (listed per AGGREGATOR)
>
> The summary stats are copied here, but please see other details at the link provided.
>
> AS_SET ANALYSIS (OREGON 2019-10-03:00)
> The Routeviews OREGON collector peers with 43 ASes.
>
> Summary stats:
>
>         Total # Updates : 30052331
>         # Updates with AS_SET : 14348
>         Percentage of Updates with AS_SET : 0.048%
>         # Total # ASes (globally) : 66205
>         # ASes that create Updates with AS_SET : 144
>         % ASes that create Updates with AS_SET : 0.218%
>
>         # Routes with AS_SET (after eliminating AS path redundancy): 477
>         Explanation: These are routes with unique {prefix, AS_SET, 1st AS after AS_SET, AGGREGATOR} combinations.
>
> Out of the 477 routes with AS_SETs:
>         *** Identifying Routes with AS_SET that seem meaningless or malformed ***
>         # Routes with only one AS in AS_SET : 383
>         # Routes with Reserved ASN in AS_SET : 131
>         # Routes with common AS in the AS_SEQUENCE and AS_SET : 139
>         # Routes with repeated ASes in the AS_SET : 0
>         # Routes that are /24 prefix (aggregate) announcements : 239
>         Total # Routes that seem meaningless or malformed : 456
>
>         Total # Routes that seem meaningful : 21
>
>         Distribution of # unique ASes in the AS_SET : 1:383, 2:68, 3:14, 4:5, 5:2, 6:3, 23:1, 31:1
>
>         # Routes with AS_SET where AGGREGATOR does not match the right most AS in AS_SEQUENCE : 47
>
>         # Routes with unique {prefix, AS_SET, AGGREGATOR} : 469
>         # Routes with unique {prefix, AS_SET} : 455
>
> *** When there is an AGGREGATOR but no AS_SET ***
>         # Unique prefixes (with or without AS_SET) : 826535
>         # Unique prefixes without AS_SET but with AGGREGATOR: 75698
>         % Unique prefixes without AS_SET but with AGGREGATOR: 9.158%
>
> Sriram



-- 
I don't think the execution is relevant when it was obviously a bad
idea in the first place.
This is like putting rabid weasels in your pants, and later expressing
regret at having chosen those particular rabid weasels and that pair
of pants.
   ---maf