Ask HN: How does Starlink work for IP address/ASN internationally?
I was curious how Starlink would work when connected in international locations? Would it show a US IP address? Would it show a Starlink ISP ASN?
For comparision - cell phone tethering plans, you can be "anywhere" in the world, but the roaming telcom provider will (lack of a better word) "tunnel" your connection to your local telcom so you could be in China, France, UK but you would have the country of origin IP address/telcom provider ASN? E.g. T-Mobile from USA, but in Europe would show that you're either in California or Kansas.
66 comments
[ 5.4 ms ] story [ 128 ms ] threadhttps://www.reddit.com/r/Starlink/comments/jm8iz9/turns_out_...
[1]: https://bgp.he.net/AS14593 [2]: https://bgp.he.net/AS14593#_prefixes [3]: https://bgp.he.net/AS14593#_prefixes6
Tada. You've got an ASN and are running it.
For example, the in-tesla car map uses various google data sources in ways the regular API doesn't support.
Most teams keep the sales guys a long way from engineering, and one customers request rarely makes it in.
I think that's because they want every user to have a reasonable experience with all the websites out there that do geoIP lookups. It also gives them flexibility in the future to not act as a worldwide internet backbone for their own users traffic by advertising each set of local IP's only at local POP's.
[1]: https://starlink-enterprise-guide.readme.io/docs/dhcp-config....
They'd need multiple ASNs for each of these network islands then, though – is their network actually set up like that?
> if you roam too far, your connection will be briefly interrupted while they assign you a new IP address.
That's an interesting detail: It seems like they might not actually have cross base station roaming/connectivity, unlike e.g. mobile ISPs in large countries (where you can keep the same IP and maintain open connections even as you're driving across the country).
I wonder if Starlink will eventually build out a terrestrial backbone to complement their satellite network for these two scenarios (i.e. very mobile base stations, such as those on airplanes, and terminating their own traffic globally as opposed to locally).
No they wouldn’t. They could connect to local transit/IXs in each region and just advertise the regional networks.
This is actually what they do when you enable Global Accelerator which is a paid feature. In GCP they have similar concept called “Premium Tier”
But basic routing entity in BGP is still prefix/route, and different routes of one AS can be announced in different ways.
I then have 2.2.3.0/24 advertised out of Frankfurt with the same AS. Again I have transit. I can reach 1.1.1.1.
How does a packet from 2.2.2.1 reach 2.2.3.1 -— 2.2.3.0 won’t be advertised by the transit to me in NY as it thinks I’ve already got the route, after all it originated from my AS.
The transit may not send you your routes with your ASN due to how its policies are configured, but that is configurable, not protocol-level issue.
And most BGP implementations have knobs to disable the loop check, so on your side you could just disable it.
That said, you can often still communicate between two NATs using various traversal techniques, although that's generally less likely for carrier-grade NATs, so NAT traversal solutions usually need relays as a fallback anyway.
GeoIP is a further complication not really related to the Internet. The DNS server in Houston can serve replies for packets coming from anywhere in the country; instead of looking at what network interface they arrived on, it looks in a database, decides 1.2.3.4 is in Oklahoma, and provides IP addresses for the Houston datacenter because of that. For Starlink to provide the best path in that case, it needs to understand what that database looks like; it's likely they populate it themselves in a protocol outside from any Internet machinery. (The most naive databases just look at the ASN of the source address, look at where the address of the ISP is, and decides that's the location of the IP address. But that is too naive to work, so nobody does that exactly.)
I would imagine that most of the CDNs / edge compute providers use some combination of these two data points to route traffic. I have never set anything like this up, so I'm not sure what the state of the art is. But, there are many tools that you can use, and none of them require having an extra ASN.
When I worked at an ISP we did some somewhat-sketchy management around this. The databases often have a flag for "is this address a cloud provider or a residential ISP subscriber?" We would see which of our IP ranges were considered residential (we had 0 residential customers), and then use them for IP transit subscribers that wanted to sell services like scraping Amazon or whatever. Amazon would check their database, say "oh this is just a virus-infected cable modem", and allow the traffic. If the reputation changed, we could just give the address to a transit customer that didn't want to do that kind of stuff. Just going to say, I played no part in this business decision; it existed when I arrived and existed when I left. Gotta pay back the investors somehow, I guess!
(At my next job I was on the other side of this; trying to combat free trial abuse. The databases were largely useless; what happens is that people at high-reputation ISPs get their servers hacked and the attacks end up launched from there. The big cloud providers were like 100% abuse traffic, probably because of free trials of CI systems and things like that. Ultimately, most traffic going through a US-based transit provider got "please contact sales to do a free trial". Actual residential ISPs like Comcast were totally fine; thanks Windows Update! Since it was the pandemic, everyone was working from home, and thus the business ISPs being gated behind "talk to sales" were not a big problem. Oh and BTW, I am certain that most of the abuse was from HN readers; it started the day after we mentioned our product here. That cat and mouse game was some of the most fun I've had at work!)
I think their stated plan is to use laser-links to satellites in higher orbits to do that. Don't know if they'll be bandwidth limited, but there's better latency doing it this way. (light goes faster in the vacuum of space, vs going down an optical fibre)
https://geoip.starlinkisp.net/feed.csv
Folks interested in geofencing content and such pull from there.
There’s also a substantial difference in performance depending on where you are, I assume depending on how saturated the local ground station is - in Portugal, I get near 200Mbps, in Spain, 140, in France, 90.
[0] https://geoip.starlinkisp.net/feed.csv
https://brr.fyi/ is a really good blog from a person nearing the end of a year long deployment who writes much more eloquently than I ever could if you want to read more.
Antarctica: A Year on Ice is an enjoyable film documentary from ~10 years ago.
A Big Dead Place by Nicholas Johnson is also worth a read from the early 2000s.
At South Pole, I believe they don't yet have access to starlink. They also don't have 24 hour coverage for the few internet satellites they do get.
I don't work in IT so don't have direct access to more specific information, and obviously I can't speak for the many other stations scattered throughout the continent.
+1, I came here out of curiosity to see what others have to say, but ended up learning something new about how roaming works.
And avoiding local ISPs may be major reason to use Starlink!
For V2 satellites added sat-sat link, but I have not hear, how large share V2 now, so I think, most links now run in compatible mode.
Other thing, for very long time, AS are classless, mean, they could be registered to any Earth GPS coordinate, but could routed via any peer.
For example, few years ago happen issue, Rostelecom announced ASNs of 90% of EU providers and for few hours, 90% of EU traffic directed via Russian links.
Etc, etc..