[Starlink] APNIC56 last week

Dave Taht dave.taht at gmail.com
Sat Sep 23 22:46:59 EDT 2023


I long for an update from mark. It was a great simulation. I would
love to see it use the known downlink data we are seeing nowadays, and
to see it calculate best case paths.

As for BARD and chatgpt: Chatgpt said I had died in 2020, "leaving
behind a substantial body of work. Bard has me presiding over
Magareatha.

It would be funnier, I suppose, if my grip on reality was weaker, or I
was less unsure I really was living in a simulation.

On Sat, Sep 23, 2023 at 2:55 PM Larry Press via Starlink
<starlink at lists.bufferbloat.net> wrote:
>
> In his first Starlink simulation, Mark Hadley assumed five transponders per satellite -- Two in-plane, two adjacent, and one crossing:
> https://www.youtube.com/watch?v=3479tkagiNo&list=PLrRMhEONgz06oMXQOljB5BoZxZw6cShLN&index=3
>
> In his next Starlink simulation, Mark Hadley assumed four:
> https://www.youtube.com/watch?v=QEIUdMiColU&list=PLrRMhEONgz06oMXQOljB5BoZxZw6cShLN&index=2
>
> I guess he assumed the relative velocities were too high for the crossing connection.
>
> I asked Bard "How many laser transponders are in a second-generation Starlink satellite?" and it said "four."
> I asked ChatGPT "How many laser transponders are in a second-generation Starlink satellite?" and it said it did not know.
> I asked Bard again and the second time it said it did not know.
> When I pointed out that it had told me "four" the first time I asked, Bard apologized for its previous answer.
>
> I remain an "AI" skeptic:
> https://circleid.com/posts/20230721-google-bard-fails-to-answer-satellite-internet-questions
>
> From: Starlink <starlink-bounces at lists.bufferbloat.net> on behalf of David Lang via Starlink <starlink at lists.bufferbloat.net>
> Sent: Friday, September 22, 2023 1:41 AM
> To: Alexandre Petrescu <alexandre.petrescu at gmail.com>
> Cc: starlink at lists.bufferbloat.net <starlink at lists.bufferbloat.net>
> Subject: Re: [Starlink] APNIC56 last week
>
> I believe that I read that STarlink has 5 lasers per sat. but whatever the
> number, it's a tiny number compared to the number of satellites that they have
> up there.
>
> As you are looking at 'trains', check their altitude. They aren't going to
> shuffle sats around much, it's expensive in terms of fuel and they are only
> allowed to provide service when they are in their proper orbits.
>
> We know the lasers are in operation as they are providing service to places more
> than one sat hop away from ground stations. We also know they have a lot of
> ground stations around to share the load.
>
> We have almost no details on the specific modules they are using, and none on
> what routing they are using.
>
> David Lang
>
>
>   On Fri, 22 Sep 2023,
> Alexandre Petrescu via Starlink wrote:
>
> > Date: Fri, 22 Sep 2023 10:26:26 +0200
> > From: Alexandre Petrescu via Starlink <starlink at lists.bufferbloat.net>
> > Reply-To: Alexandre Petrescu <alexandre.petrescu at gmail.com>
> > To: starlink at lists.bufferbloat.net
> > Subject: Re: [Starlink] APNIC56 last week
> >
> >
> > Le 21/09/2023 à 21:05, Inemesit Affia via Starlink a écrit :
> >> Not going to go into details but lasers have been identified in photos
> >> of the sats and one of the component suppliers is known. (The scale is
> >> novel, not the tech, demisabiliy is new though)
> >
> > 4 or 2 lasers on each sat  (N-S, E-W) is potentially a very different
> > matter from an IP routing standpoint.  It still is a reduced set of
> > variables, for a routing protocol (it is not like there being an
> > arbitrary number of IP interfaces, it's just 2 or 4).
> >
> > For component manufacturers: yes, I heard about a few manufacturers of
> > such equipment for laser comms for LEO sats, experimented.  There is
> > public information about a few of them.  I dont know which is considered
> > by starlink, but there is not my worry.  There is also a difference
> > between laser links between sats on different orbit altitudes (e.g.
> > laser for ISL for GEO to MEO) and lasers between sats on a same orbit
> > altitude, or on a same orbit.  It's three different things, with
> > different sets of requirements: focusing, power levels, distance ranges.
> >
> > At the lowest limit (cheapest, less powerful, less range distance), I
> > suppose it is possible to use simply LiFi optical links (a sort of WiFi
> > but with light).  If so, then it is very easy to have IP on it.
> >
> > There is also an 'optical' spec that was circulated here on this list
> > (https://urldefense.com/v3/__https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf__;!!P7nkOOY!pso4rogBStK7jR7pEg21ACCfJyL_ZCltZip5K7pxmJhmaSKfOY1RgwXzW4EtPOvEhS9XoLRaKS8QbVoEgYhXDWmzfQ$ ),
> > although it seemed to me to figure on kepler's website, not on
> > starlink's.  In that spec, it is said Ethernet, among other things.  On
> > Ethernet, IP can run easily.
> >
> >>
> >> Starlink can't deliver to Antarctica or Northern parts of Alaska,
> >> Ascension Island, Diego Garcia, Easter Island, Vanuatu, Iran without
> >> ISL's etc
> >
> > I'll have to look where these places are.
> >
> > When looking at starlink satellites I often see trains forming and
> > lasting for a while.  Someone said these sats are like that (trains)
> > prior to be put on a more evenly distanced, in-orbit; but some time
> > passed, and they continue this kind of behaviour: form denser trains,
> > then distance more evenly, and back again.  So I am not sure these
> > 'trains' are ephemeral.  They seem to be in such 'train' structure while
> > above some particular continents or areas, but not sure.  It takes a lot
> > of time to make a meaning of it.
> >
> > Also, now here are at least two kinds of starlink subscription plans:
> > 40EUR/month and 287EUR/month, for fixed vs mobile.
> >
> >>
> >> North South links seem to work but not East West (if they exist)
> >
> > Yes, good question.  It makes a lot of difference whether there are 2 or
> > 4 laser links on each sat.  It also makes a lot of difference if trying
> > to make IP routing work there (assuming there could be 2 or 4 IP
> > interfaces for lasers).
> >
> > This (number of ISL links on a starlink sat) can have an impact on how
> > people show LEO satellite topologies in Internet Drafts at IETF.
> >
> > Alex
> >
> >>
> >> On Thu, Sep 21, 2023, 2:20 PM Alexandre Petrescu via Starlink
> >> <starlink at lists.bufferbloat.net> wrote:
> >>
> >>
> >>     Le 19/09/2023 à 06:39, Ulrich Speidel via Starlink a écrit :
> >>     > FWIW, I gave a talk about Starlink - insights from a year in -
> >>     at last
> >>     > week's APNIC56 conference in Kyoto:
> >>     >
> >>     > https://urldefense.com/v3/__https://conference.apnic.net/56/program/program/*/day/6/technical-2/__;Iw!!P7nkOOY!pso4rogBStK7jR7pEg21ACCfJyL_ZCltZip5K7pxmJhmaSKfOY1RgwXzW4EtPOvEhS9XoLRaKS8QbVoEgYjUChNXHQ$
> >>
> >>
> >>     Thanks for the presentation.
> >>
> >>     I would like to ask what do you mean by "Method #2: "space
> >>     lasers""and
> >>     "Not all Starlink satellites have
> >>     lasers" on slide 5?
> >>
> >>     It seems to be saying there is inter-satellite communications. The
> >>     need
> >>     of that seems to stem from the lack of ground 'teleport' that is
> >>     necessary for DISHY-SAT-Internet communications, so a SAT-to-SAT
> >>     communication is apparently used with lasers.  I can agree with
> >>     the need.
> >>
> >>     What standard is used for these lasers?
> >>
> >>     Is this ISL communicaiton within the starlink constellation a
> >>     supposition or a sure thing?
> >>
> >>     Other presentations of starlink mentioned on this list dont talk
> >>     about
> >>     this lasers between sats (dont show  lasers on the sats), but kepler
> >>     talks about optical links, and also there is talk about ISOC LEO
> >>     Internet about such 'lasers from space'.
> >>
> >>     (I must say that I thought previously that there were only 2 or 3
> >>     ground
> >>     teleports overall in EU and USA, but I see now there is a teleport
> >>     in NZ
> >>     too).
> >>
> >>     (for price comparison: it is said 100USD monthly, but in France right
> >>     now the monthly subscription is at around 40 Euros;  this competes
> >>     very
> >>     advantageously to other satcoms ISPs for rural areas non-covered
> >>     by 5G;
> >>     the cellular monthly subscriptions are still much more advantageous,
> >>     where there is 5G, of course).
> >>
> >>     Alex
> >>
> >>     >
> >>     > Also well worth looking at is Geoff Huston's excellent piece on the
> >>     > foreseeable demise of TCP in favour of QUIC in the same session.
> >>     One
> >>     > of Geoff's main arguments is that the Internet is becoming local,
> >>     > i.e., most traffic goes between a CDN server and you, and most
> >>     data is
> >>     > becoming proprietary to the application owner, meaning it suits the
> >>     > Googles and Facebooks of this world very well not to be using
> >>     TCP for
> >>     > its transport, but rather pull the transport specifics into the
> >>     > application layer where the have full control.
> >>     >
> >>     > Food for thought, especially since LEO networks are a
> >>     particularly bad
> >>     > place to put local content caches, since the concept of what's
> >>     "local"
> >>     > in a LEO network changes constantly, at around 20,000 miles an
> >>     hour or
> >>     > so. Spoke to a Rwandan colleague who installs Starlink there and
> >>     sees
> >>     > all traffic to anywhere go via the US with RTTs of nearly 2
> >>     seconds,
> >>     > even if the Rwandan user is trying to access a Rwandan service.
> >>     >
> >>     > About to hop onto a plane (ZK-NZJ) tonight with free WiFi (Ka band
> >>     > GEO) enroute to Auckland in the hope of getting a better experience
> >>     > than last time when the system seemed to run out of IP addresses on
> >>     > its DHCP.
> >>     >
> >>     _______________________________________________
> >>     Starlink mailing list
> >>     Starlink at lists.bufferbloat.net
> >>     https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/starlink__;!!P7nkOOY!pso4rogBStK7jR7pEg21ACCfJyL_ZCltZip5K7pxmJhmaSKfOY1RgwXzW4EtPOvEhS9XoLRaKS8QbVoEgYgJ4na9hA$
> >>
> >>
> >> _______________________________________________
> >> Starlink mailing list
> >> Starlink at lists.bufferbloat.net
> >> https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/starlink__;!!P7nkOOY!pso4rogBStK7jR7pEg21ACCfJyL_ZCltZip5K7pxmJhmaSKfOY1RgwXzW4EtPOvEhS9XoLRaKS8QbVoEgYgJ4na9hA$
> > _______________________________________________
> > Starlink mailing list
> > Starlink at lists.bufferbloat.net
> > https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/starlink__;!!P7nkOOY!pso4rogBStK7jR7pEg21ACCfJyL_ZCltZip5K7pxmJhmaSKfOY1RgwXzW4EtPOvEhS9XoLRaKS8QbVoEgYgJ4na9hA$
> >
> _______________________________________________
> Starlink mailing list
> Starlink at lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/starlink



-- 
Oct 30: https://netdevconf.info/0x17/news/the-maestro-and-the-music-bof.html
Dave Täht CSO, LibreQos
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 84813 bytes
Desc: not available
URL: <https://lists.bufferbloat.net/pipermail/starlink/attachments/20230923/8172e815/attachment-0001.png>


More information about the Starlink mailing list