[Starlink] a puzzling starlink uplink trace

Sebastian Moeller moeller0 at gmx.de
Wed Aug 30 15:43:08 EDT 2023


Hi Dave,


you probably know already, but
mtr -ezb4 www.heise.de
will report compliant MPLS segments:

                                                                                          My traceroute  [v0.95]
123-1234567.local (192.168.42.229) -> www.heise.de (193.99.144.85)                                                                                                               2023-08-30T21:39:48+0200
Keys:  Help   Display mode   Restart statistics   Order of fields   quit
                                                                                                                                                                 Packets               Pings
 Host                                                                                                                                                          Loss%   Snt   Last   Avg  Best  Wrst StDev
 1. AS???    192.168.42.1 (192.168.42.1)                                                                                                                        0.0%  1067    0.9   0.6   0.5   9.1   0.4
 2. AS6805   loopback1.0003.acln.06.ham.de.net.telefonica.de (62.52.201.200)                                                                                    0.0%  1067   12.3  22.5   9.9 172.7  17.5
 3. AS6805   bundle-ether10.0002.dbrx.06.ham.de.net.telefonica.de (62.53.2.98)                                                                                  0.7%  1067   12.4  12.0  10.3  16.5   0.7
 4. AS6805   ae7-0.0001.corx.06.ham.de.net.telefonica.de (62.53.15.0)                                                                                           0.0%  1067   18.7  19.2  17.2  48.9   2.4
    [MPLS: Lbl 16624 TC 0 S 1 TTL 1]
 5. AS6805   ae6-0.0002.corx.02.fra.de.net.telefonica.de (62.53.0.49)                                                                                           0.0%  1067   41.6  19.6  17.2  51.6   3.9
    [MPLS: Lbl 16624 TC 0 S 1 TTL 1]
 6. AS6805   bundle-ether2.0001.cord.01.off.de.net.telefonica.de (62.53.0.199)                                                                                  0.0%  1067   19.0  19.3  17.7  87.1   2.2
    [MPLS: Lbl 16624 TC 0 S 1 TTL 1]
 7. AS6805   bundle-ether1.0002.corp.01.off.de.net.telefonica.de (62.53.28.171)                                                                                 0.0%  1067   22.9  19.2  17.5  29.7   0.9
 8. AS???    ipv4.de-cix.fra.de.as12306.plusline.net (80.81.192.132)                                                                                            0.0%  1066   18.5  19.7  18.2  93.0   2.5
 9. AS12306  82.98.102.71 (82.98.102.71)                                                                                                                        0.0%  1066   19.6  19.4  17.7  69.5   1.7
    [MPLS: Lbl 24002 TC 0 S 1 TTL 1]
10. AS12306  82.98.102.23 (82.98.102.23)                                                                                                                        0.0%  1066   17.8  19.8  17.4 197.7   9.9
11. AS12306  212.19.61.13 (212.19.61.13)                                                                                                                        0.0%  1066   19.8  19.6  17.5 155.7   6.1
12. AS12306  www.heise.de (193.99.144.85)                                                                                                                       0.1%  1066   19.6  19.1  17.6  63.7   1.5

Not that these do not already stick out as hops at different locations (here Hamburg (ham), Frankfurt (fra) and Offenbach)) all with more or less identical distance-independent RTTs.

Other than that quite intriguing puzzle, I wonder whether we could teach flent to run a bidirectional traceroute/mtr as part of its measurements..?

Regards
	Sebastian


> On Aug 30, 2023, at 20:07, Dave Taht via Starlink <starlink at lists.bufferbloat.net> wrote:
> 
> In the attached 5 minute plot from a few days ago (I can supply the
> flent.gz files if anyone wants them), I see a puzzling spike at T+155s
> to nearly 90ms of baseline latency, then down to 20ms. No degree of
> orbital mechanics can apply to this change, even factoring in an over
> the horizon connection, routing packets on the ground through LA to
> seattle, and back, or using a couple ISLs, can make this add up for
> me. A combination of all that, kind of does make sense.
> 
> The trace otherwise shows the sawtooth pattern of a single tcp flow ,
> a loss (sometimes catastrophic) at every downward bandwidth change.
> 
> An assumption I have long been making is the latency staircase effect
> (see T+170) forward is achieving the best encoding rate at the
> distance then seen, the distance growing and the encoding rate falling
> in distinct steps, with a fixed amount of buffering, until finally
> that sat starts falling out of range, and it choses another at T+240s.
> 
> But jeeze, a 70ms baseline latency swing? What gives?  I imagine
> somehow correlating this with a mpls enabled traceroute might begin to
> make some sense of it, correlated by orbital positions....
> 
> 
> -- 
> Oct 30: https://netdevconf.info/0x17/news/the-maestro-and-the-music-bof.html
> Dave Täht CSO, LibreQos
> <thestarlinkmystery.png>_______________________________________________
> Starlink mailing list
> Starlink at lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/starlink



More information about the Starlink mailing list