[afnog] AMS-IX concerns

[Frank Habicht]

Hi Adam,

according to
https://eastafrica.ams-ix.net/connected_parties
there are 4 peers connected (IXP NOC not counted).

if one or two of them occasionally see better routes (not via IXP) to
destinations at one or two of the other peers, something like this can
easily happen imho.

It look like it happens quite often, and doesn't last long, so routes are
unstable. Generally not good. that happening only during work hours I would
guess one (or more) humans are part of this - ie not because of saturation.
My _feeling_ ....

only participating networks, or route server operator are really able to tell.
(Oh, and I heard: one of the participants is doing a network upgrade)

Frank


On 10/29/2014 6:03 PM, Adam Nelson wrote:
> Hello,
> 
> I'm looking at the following chart:
> 
> https://eastafrica.ams-ix.net/technical--4/statistics
> https://dl.dropboxusercontent.com/s/ncjo6o4yi0lgivd/2014-10-29%20at%202.24%20PM.png?dl=0
> 
> I've looked at enough charts to see that the flat top at 1Gbps with deep
> dips signals a saturation of some sort.
> 
> However, I'm getting a very negative response from AMS-IX and since they
> won't give out peering data, there's no way to confirm my suspicions that
> something is wrong there.
> 
> Anyway, does anybody have any information to support or refute the claim? 
> I may be totally wrong (which would be fine) but I really don't know. 
> Unfortunately, JTL, Safaricom, and KENET are all mostly down today so I'm
> having trouble doing much research.
> 
> Cheers,
> Adam
> --
> Kili - Cloud for Africa: kili.io <http://kili.io/>
> Musings: twitter.com/varud <https://twitter.com/varud>
> More Musings: varud.com <http://varud.com>
> About Adam: www.linkedin.com/in/adamcnelson
> <https://www.linkedin.com/in/adamcnelson>
> 
> 
> _______________________________________________
> afnog mailing list
> http://afnog.org/mailman/listinfo/afnog
>