Starlink has bufferbloat. Bad.
 help / color / mirror / Atom feed
From: Inemesit Affia <inemesitaffia@gmail.com>
To: j pan <panatuvicdotca@gmail.com>,
	Dave Taht via Starlink <starlink@lists.bufferbloat.net>
Subject: [Starlink] Re: New Customer or route leak?
Date: Sun, 5 Jul 2026 08:06:53 +0100	[thread overview]
Message-ID: <813df236-6fe7-4626-b86e-ad22584eca2a@gmail.com> (raw)
In-Reply-To: <CAHn=e4hygFqP-ksdm3DSmeCgEPdfVJS3kSXVu7KE1P44DA23Ag@mail.gmail.com>

I don't have a shell or VPS. I'm using Globalping's website via a remotely accessed browser. (My environment is very locked down sorry)

Since AS10075 is in Starlink's AS-SET we know it's at least somewhat deliberate i.e it isn't a route leak. (Similar to what Vocus and Telstra do)

2403:cd40:8::/48 also has a route that may help give you another vantage point

10075 14593 55850 6461

I'll see what I can do later without spending money on a VPS. GSL has a ping testing interface on their fake looking glass

Jul 5, 2026 1:49:49 AM j pan <panatuvicdotca@gmail.com>:

> it's not clear whether a community gateway is involved or traceroute
> just transits through starlink's ground backbone. can you run "ping -D
> -i 0.01 -c 10000 103.179.62.17 > out.txt" on vultr sydney or better
> singapore in addition to "traceroute -AIe 103.179.62.17" and share the
> results with me? cheers.  -j
> -- 
> J Pan, UVic CSc, ECS566, 250-472-5796 (NO VM), Pan@UVic.CA, Web.UVic.CA/~pan
> 
> On Fri, Jul 3, 2026 at 9:50 AM Inemesit Affia <inemesitaffia@gmail.com> wrote:
>> 
>> Attached is a better testing pair
>> 
>> Jul 3, 2026 5:10:20 PM Inemesit Affia <inemesitaffia@gmail.com>:
>> 
>> Try putting that prefix  in BGP.tools or bgp.he and you'll see there's no equal distance route to GSL from that prefix. 103.179.62.0/24
>> 
>> 
>> The AS path you'll see is 10075 14593 137408
>> 
>> You can try an Australian/Frankfurt/Ashburn looking Glass from GSL or Vultr in Sydney.
>> 
>> 
>> 
>> Maybe even Vocus or Lumen in Asia/Oceania
>> 
>> From Vultr Sydney
>> 
>> 45.76.118.1[Vultr]
>> ???
>> 100.100.200.1
>> 10.74.x.x.
>> 10.74.x.x.
>> 202.77.88.23
>> 206.148.24.194
>> 206.148.24.194
>> 206.148.x.x
>> 206.148.x.x
>> 160.202.164.99[GSL]
>> 206.224.74.121 [Starlink backbone]
>> ???
>> 206.214.228.69 [SB]
>> 163.47.158.229[10075]
>> 103.179.62.1[10075]
>> 
>> While I looked into it I found out it looks like an asymmetric route.
>> 
>> 
>> Attached is the traceroute for the forward and reverse path
>> 
>> Jul 3, 2026 3:40:26 PM j pan <panatuvicdotca@gmail.com>:
>> 
>> how do you know it traverses through starlink? it does not appear so
>> from the following
>> 
>> traceroute to 1.1.1.1 (1.1.1.1), 20 hops max, 60 byte packets
>> 1  _gateway (103.131.159.193)  0.645 ms  0.617 ms
>> 2  103.179.62.76 (103.179.62.76)  0.598 ms  0.596 ms
>> 3  103.131.159.145 (103.131.159.145)  1.257 ms  1.261 ms
>> 4  103.7.249.234 (103.7.249.234)  26.821 ms  26.818 ms
>> 5  one.one.one.one (1.1.1.1)  1.070 ms  1.067 ms
>> -- 
>> J Pan, UVic CSc, ECS566, 250-472-5796 (NO VM), Pan@UVic.CA, Web.UVic.CA/~pan
>> 
>> On Fri, Jul 3, 2026 at 4:14 AM Inemesit Affia via Starlink
>> <starlink@lists.bufferbloat.net> wrote:
>> 
>> 
>> 103.179.62.0/24 and other prefixes from a Bangladeshi provider are going to Australia(Singapore?) {GSL} via Starlink.
>> 
>> I'm not seeing any latency improvements (HE has similar performance in Sydney) and I also noticed Vocus is taking some traffic from Singapore via Perth to Bangladesh via this path.
>> 
>> Seen probes in Frankfurt/Ashburn/elsewhere do the similar.
>> 
>> 
>> Is there really a Community Gateway working this?
>> 
>> Where's it(Singapore/Australia)?
>> 
>> When I started at a large multinational Telco decades ago, I met a system that monitored/tracked path latency to pretty much every country/provider and alerted on shifts, is that something unusual to have?
>> _______________________________________________
>> Starlink mailing list -- starlink@lists.bufferbloat.net
>> To unsubscribe send an email to starlink-leave@lists.bufferbloat.net

  parent reply	other threads:[~2026-07-05  7:07 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-03 11:14 [Starlink] New Customer or route leak? Inemesit Affia
2026-07-03 14:40 ` [Starlink] " j pan
2026-07-03 16:09   ` Inemesit Affia
2026-07-03 16:49     ` Inemesit Affia
     [not found]       ` <CAHn=e4hygFqP-ksdm3DSmeCgEPdfVJS3kSXVu7KE1P44DA23Ag@mail.gmail.com>
2026-07-05  7:06         ` Inemesit Affia [this message]
2026-07-05 16:02           ` Inemesit Affia
2026-07-05 16:38             ` Inemesit Affia
2026-07-06  6:57               ` Jianping Pan
2026-07-03 17:36     ` Jianping Pan
2026-07-03 18:03       ` Inemesit Affia
2026-07-04 23:41         ` Jianping Pan

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=813df236-6fe7-4626-b86e-ad22584eca2a@gmail.com \
    --to=inemesitaffia@gmail.com \
    --cc=panatuvicdotca@gmail.com \
    --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