Starlink has bufferbloat. Bad.
 help / color / mirror / Atom feed
From: Vint Cerf <vint@google.com>
To: Dave Taht <dave.taht@gmail.com>
Cc: Nathan Owens <nathan@nathan.io>,
	Starlink list <starlink@lists.bufferbloat.net>
Subject: Re: [Starlink] Starlink no longer available to the Ukrainian army?
Date: Sun, 16 Oct 2022 14:29:20 -0400	[thread overview]
Message-ID: <CAHxHggeZHQmeHfkhRGfMyay8HRDLfKPcAh4oUJ5dANoWBHbbig@mail.gmail.com> (raw)
In-Reply-To: <CAA93jw7WGW7JZbkYDdkZR43_9AxHpJj=oBTyjozLhOYKbY6rAg@mail.gmail.com>


[-- Attachment #1.1: Type: text/plain, Size: 7419 bytes --]

if you need real-time for video and control, that can add up ...

v


On Sun, Oct 16, 2022 at 2:01 PM Dave Taht via Starlink <
starlink@lists.bufferbloat.net> wrote:

> On Sun, Oct 16, 2022 at 10:57 AM Nathan Owens via Starlink
> <starlink@lists.bufferbloat.net> wrote:
> >
> > Elon posted a graph, it showed a peak of 7000GB per unit time, the only
> one that makes sense to me is per hour, which is 15Gbps peak -- not a huge
> amount.
>
> You really don't need much data for C&C traffic. You do need a fairly
> reliable uplink, but the observed jitter on an otherwise idle link was
> in the few ms range.
>
> GPSD has a udp output mode, too.
>
> (btw, to my knowledge, starlink has not enabled any network interfaces
> to the outside to their internal on-dish gps chip, which
>  when I was whinging about it, would provide perfect time to
> downstream clients, either natively or via ntp)
>
> It's one very short message, per second.
>
> >
> > On Sun, Oct 16, 2022 at 10:50 AM Steve Stroh via Starlink <
> starlink@lists.bufferbloat.net> wrote:
> >>
> >> I’m speculating, but given that Starlink is THE communications
> infrastructure for much of Ukraine, then the scaling of the ground stations
> to provide that level of service must be a significant expense. To provide
> that much bandwidth would require deploying a lot of ground stations, each
> with expensive hardware, power infrastructure (including backup), fiber
> backhaul, skilled labor, and no small amount of fiber bandwidth that SpaceX
> has to pay SOMEONE to provide.
> >>
> >> Not to mention that anything SpaceX deploys to support Ukraine is a
> resource that it could have used for speeding up revenue generation in
> lucrative markets like the US.
> >>
> >> On Fri, Oct 14, 2022 at 12:41 David Lang via Starlink <
> starlink@lists.bufferbloat.net> wrote:
> >>>
> >>> If spacex is providing the high-end/business grade service to all
> terminals
> >>> that they normally charge $4500/month for, reimbursement should be
> based on
> >>> that.
> >>>
> >>> Base it on the normal service pricing, not on cost-plus (if it were
> based on
> >>> cost-plus it would be an utter windfall for SpaceX as they are still
> in the
> >>> stage of building the service, and so there is a much higher spend
> rate to
> >>> expand the service at this point than the ongoing maintinance of it)
> >>>
> >>> while the satellites do support that area, they also support the rest
> of the
> >>> service, and if they weren't supporting Ukraine, there wouldn't be any
> fewer
> >>> satellites launched.
> >>>
> >>> I've seen too many games played with 'fully loaded costs' (sometimes
> backfiring
> >>> on the people tinkering with the numbers), and so it's something I
> watch out
> >>> for.
> >>>
> >>> lies, damn lies, and statistics, 'fully loaded costs' tend to be heavy
> on
> >>> statistics ;-)
> >>>
> >>> David Lang
> >>>
> >>> On Fri, 14 Oct 2022, tom@evslin.com wrote:
> >>>
> >>> > Putting aside the timing of Elon's complaint about cost right after
> the spat over his Ukrainian "peace plan", It is certainly reasonable for
> Starlink to get paid like other weapon suppliers who didn't give out free
> samples to prove their usefulness, Given that they should be reimbursed
> based on loaded cost plus profit like anyone else. I'm sure the other
> suppliers allocate their overhead costs when pricing weapon systems. They'd
> be out of business otherwise. The satellites are part of Starlink's fixed
> overhead so a portion of their costs should be allocated to service
> provided in Ukraine.
> >>> >
> >>> > All that being said, it would be terrible if Ukraine got less than
> the best support that can be provided.
> >>> >
> >>> > -----Original Message-----
> >>> > From: Starlink <starlink-bounces@lists.bufferbloat.net> On Behalf
> Of David Lang via Starlink
> >>> > Sent: Friday, October 14, 2022 1:28 PM
> >>> > To: Kurtis Heimerl <kheimerl@cs.washington.edu>
> >>> > Cc: Starlink list <starlink@lists.bufferbloat.net>
> >>> > Subject: Re: [Starlink] Starlink no longer available to the
> Ukrainian army?
> >>> >
> >>> > Having now read more info on this, less significant than the $80m
> total figure is the $20m/month figure he quoted. With 15k dishes as the
> figure that they sent (separate from whatever has been purchased on the
> commercial side), that works out to 1.3k/dish/month, which is very high.
> >>> >
> >>> > now, not being able to deploy reliable ground stations inside
> Ukraine could be driving up costs, plus the ongoing battle against jamming.
> But in his tweet he also cites satellite costs, which should not be
> allocated as "Ukraine related"
> >>> > costs (and I don't think the cyberdefense and jamming defense work
> should be
> >>> > either)
> >>> >
> >>> > David Lang
> >>> >
> >>> > On Fri, 14 Oct 2022, Kurtis Heimerl via Starlink wrote:
> >>> >
> >>> >> This thread (https://twitter.com/dim0kq/status/1580827171903635456)
> >>> >> strongly argues that Starlink is largely paid for their service, at
> >>> >> least on the consumer side. I imagine there are significant
> >>> >> operational expenses in dealing with the various actors involved but
> >>> >> not on the basic model.
> >>> >>
> >>> >> On Fri, Oct 14, 2022 at 9:06 AM Juliusz Chroboczek via Starlink
> >>> >> <starlink@lists.bufferbloat.net> wrote:
> >>> >>>
> >>> >>>> In essence, once you give something away for free, not even
> setting
> >>> >>>> the expectation that it’s a “freemium” model, it’s very hard to
> get
> >>> >>>> out of it. If you then claim your costs are way higher than what
> >>> >>>> analysis work out, eyebrows raise way above the hairline.
> >>> >>>
> >>> >>> Uh.  Hmm.
> >>> >>> _______________________________________________
> >>> >>> Starlink mailing list
> >>> >>> Starlink@lists.bufferbloat.net
> >>> >>> https://lists.bufferbloat.net/listinfo/starlink
> >>> >> _______________________________________________
> >>> >> Starlink mailing list
> >>> >> Starlink@lists.bufferbloat.net
> >>> >> https://lists.bufferbloat.net/listinfo/starlink
> >>> >
> >>> >_______________________________________________
> >>> Starlink mailing list
> >>> Starlink@lists.bufferbloat.net
> >>> https://lists.bufferbloat.net/listinfo/starlink
> >>
> >> --
> >> Steve Stroh N8GNJ (he / him / his)
> >> Editor
> >> Zero Retries Newsletter - https://zeroretries.substack.com
> >>
> >> _______________________________________________
> >> Starlink mailing list
> >> Starlink@lists.bufferbloat.net
> >> https://lists.bufferbloat.net/listinfo/starlink
> >
> > _______________________________________________
> > Starlink mailing list
> > Starlink@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/starlink
>
>
>
> --
> This song goes out to all the folk that thought Stadia would work:
>
> https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366665607352320-FXtz
> Dave Täht CEO, TekLibre, LLC
> _______________________________________________
> Starlink mailing list
> Starlink@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/starlink
>


-- 
Please send any postal/overnight deliveries to:
Vint Cerf
Google, LLC
1900 Reston Metro Plaza, 16th Floor
Reston, VA 20190
+1 (571) 213 1346


until further notice

[-- Attachment #1.2: Type: text/html, Size: 11137 bytes --]

[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 3995 bytes --]

  reply	other threads:[~2022-10-16 18:29 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-14 12:06 Juliusz Chroboczek
2022-10-14 12:45 ` David Lang
2022-10-14 13:05   ` Juliusz Chroboczek
2022-10-14 13:11     ` David Lang
2022-10-14 14:17       ` Mike Puchol
2022-10-14 18:26         ` Juliusz Chroboczek
2022-10-14 20:44           ` Mike Puchol
2022-10-16 17:31             ` Steve Stroh
2022-10-16 17:56               ` Dave Taht
2022-10-16 18:28               ` Vint Cerf
2022-10-14 12:51 ` Mike Puchol
2022-10-14 13:13   ` Juliusz Chroboczek
2022-10-14 14:15     ` Mike Puchol
2022-10-14 16:05       ` Juliusz Chroboczek
2022-10-14 17:13         ` Kurtis Heimerl
2022-10-14 17:28           ` David Lang
2022-10-14 18:16             ` Juliusz Chroboczek
2022-10-14 18:32             ` Larry Press
2022-10-14 19:32               ` Benjamin Henrion
2022-10-14 19:35               ` David Lang
2022-10-14 20:07                 ` David Lang
2022-10-14 22:29                   ` Dave Taht
2022-10-15 16:51                   ` Larry Press
2022-10-16  6:42                     ` Sebastian Moeller
2022-10-16  8:45                       ` Bruce Perens
2022-10-16 16:31                         ` Dave Taht
2022-10-16 17:00                           ` Sebastian Moeller
2022-10-16 19:55                           ` Brandon Butterworth
2022-10-16 20:25                           ` Bruce Perens
2022-11-02  3:50                             ` [Starlink] 25 years of open source - where we went wrong, and how to fix it Dave Taht
2022-10-14 19:13             ` [Starlink] Starlink no longer available to the Ukrainian army? tom
2022-10-14 19:41               ` David Lang
2022-10-14 20:09                 ` Dave Taht
2022-10-16 17:50                 ` Steve Stroh
2022-10-16 17:57                   ` Nathan Owens
2022-10-16 18:01                     ` Dave Taht
2022-10-16 18:29                       ` Vint Cerf [this message]
2022-10-16 18:33                         ` David Lang
2022-10-16 18:44                           ` Sebastian Moeller
2022-10-16 19:03                             ` Brandon Butterworth
2022-10-16 19:37                               ` Ulrich Speidel
2022-10-16 19:51                             ` David Lang
2022-10-16 20:37                               ` Michael Richardson
2022-10-16 20:52                                 ` Bruce Perens
2022-10-16 22:17                                 ` Brandon Butterworth

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://lists.bufferbloat.net/postorius/lists/starlink.lists.bufferbloat.net/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAHxHggeZHQmeHfkhRGfMyay8HRDLfKPcAh4oUJ5dANoWBHbbig@mail.gmail.com \
    --to=vint@google.com \
    --cc=dave.taht@gmail.com \
    --cc=nathan@nathan.io \
    --cc=starlink@lists.bufferbloat.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox