4 ms·

It's been 20+ years since I had reason to know the internals of BGP, but I still carry a slight nervousness whenever I think about it. In many ways, it's beaut

It's been 20+ years since I had reason to know the internals of BGP, but I still carry a slight nervousness whenever I think about it. In many ways, it's beautiful in that it allows Autonomous Systems to come together and independently make the Internet. However, I always felt it was brittle in that it relied every AS to have edge routers aware of the entire AS routing map, and so when those routers went down hard, when they came back up they were blank and needed to learn the Internet again. I noted back then that on some routers a single spoofed UDP packet to a router getting a table from its peers could cause it to stop doing that and start again. In a time when those updates could take hours... well... that kept me awake at night sometimes. Propagation of routes was global and error prone. Even today, you can see on the spy glass websites rogue AS routers advertising crazy routes either by accident or on purpose as a hostile act (to try and get traffic for a target network through itself). There are tens of thousands of engineers globally nursing and managing this stuff to keep the whole thing going. Like I said, it's been a while since I last looked closely, and I imagine multiple improvements have been made in this space since that time, but if there was ever a protocol that needs a long, cold hard look for a replacement, it's quite possibly BGP. It's also the one - and only - non-crypto situation where if somebody asked me if a blockchain could be really very useful, I'd say "probably, yeah".

nroets 5y ago
So much of the internet is about translating addresses from one format to another: Human readable to IP: DNS Private IP to Public IP: NAT MAC to Private IP: ARP Public IP to AS: BGP With a block chain DNS it may be possible to replace all of them with a single protocol, but it may well turn out to be an academic exercise.
- grlass 5y ago
  Namecoin [1] was an early fork of bitcoin, that did blockchain based DNS, with the ideas being actively discussed by Satoshi Nakamoto [2]. That deals with DNS (though how much has it been used?), but what about the layers of the stack? [1] https://en.wikipedia.org/wiki/Namecoin https://en.wikipedia.org/wiki/Namecoin [2] https://bitcointalk.org/index.php?topic=1790.222 https://bitcointalk.org/index.php?topic=1790.222
- dgb23 5y ago
  These are abstraction barriers that isolate locations, layers and protocols from each other. I don't understand this as translation but rather as a separation of concerns.
- oblio 5y ago
  > So much of the internet is about translating addresses from one format to another Most of computing is about mapping A to B and maintaining databases of those mappings :-)
  - WorldMaker 5y ago
    As the saying goes there are two hard problems in computing: naming things, cache invalidation, and buffer overflows. Most of these systems are about naming things and cache invalidation.
    - RustyConsul 5y ago
      but if it's base zero, you still wouldn't get a buffer overflow ;)
      - WorldMaker 5y ago
        Unless it needs to be a null pointer terminated array.
  - PaulRobinson 5y ago
    Thanks for not pushing back too hard on my (elderly?), ignorance, and giving me a chance to catch up with the current state of the art. This looks very interesting, and I am looking forward to spending time diving in for no more than a sense of curiosity. Thank you.
- alfalfasprout 5y ago
  Problem is the blockchain stuff usually ends up being computationally expensive for what it's trying to do. I could maybe see a use case for DNS but then either each computer has to have the full blockchain locally or rely on providers to do that for them in which case we're basically at the same model we have now (with root servers and DNS servers). Obviously you don't want to publish private IPs, etc. to other computers... so inside a local network you're not buying yourself much. For routing tables I could see more of a use case for blockchain technology since you can actually verify the routing tables and prevent BGP spoofing, etc.
- collinmanderson 5y ago
  I believe IPv6 intended to replace both ARP/MAC and NAT, but has mostly failed to achieve that.
vlan0 5y ago
>Like I said, it's been a while since I last looked closely, and I imagine multiple improvements have been made in this space since that time For anyone that's interested in what these may be. Take a peak at MANRS. https://www.manrs.org/ https://www.manrs.org/

nroets 5y ago

So much of the internet is about translating addresses from one format to another: Human readable to IP: DNS Private IP to Public IP: NAT MAC to Private IP: ARP Public IP to AS: BGP With a block chain DNS it may be possible to replace all of them with a single protocol, but it may well turn out to be an academic exercise.

grlass 5y ago
Namecoin [1] was an early fork of bitcoin, that did blockchain based DNS, with the ideas being actively discussed by Satoshi Nakamoto [2]. That deals with DNS (though how much has it been used?), but what about the layers of the stack? [1] https://en.wikipedia.org/wiki/Namecoin https://en.wikipedia.org/wiki/Namecoin [2] https://bitcointalk.org/index.php?topic=1790.222 https://bitcointalk.org/index.php?topic=1790.222
dgb23 5y ago
These are abstraction barriers that isolate locations, layers and protocols from each other. I don't understand this as translation but rather as a separation of concerns.
oblio 5y ago
> So much of the internet is about translating addresses from one format to another Most of computing is about mapping A to B and maintaining databases of those mappings :-)
- WorldMaker 5y ago
  As the saying goes there are two hard problems in computing: naming things, cache invalidation, and buffer overflows. Most of these systems are about naming things and cache invalidation.
  - RustyConsul 5y ago
    but if it's base zero, you still wouldn't get a buffer overflow ;)
    - WorldMaker 5y ago
      Unless it needs to be a null pointer terminated array.
- PaulRobinson 5y ago
  Thanks for not pushing back too hard on my (elderly?), ignorance, and giving me a chance to catch up with the current state of the art. This looks very interesting, and I am looking forward to spending time diving in for no more than a sense of curiosity. Thank you.
alfalfasprout 5y ago
Problem is the blockchain stuff usually ends up being computationally expensive for what it's trying to do. I could maybe see a use case for DNS but then either each computer has to have the full blockchain locally or rely on providers to do that for them in which case we're basically at the same model we have now (with root servers and DNS servers). Obviously you don't want to publish private IPs, etc. to other computers... so inside a local network you're not buying yourself much. For routing tables I could see more of a use case for blockchain technology since you can actually verify the routing tables and prevent BGP spoofing, etc.
collinmanderson 5y ago
I believe IPv6 intended to replace both ARP/MAC and NAT, but has mostly failed to achieve that.

vlan0 5y ago

>Like I said, it's been a while since I last looked closely, and I imagine multiple improvements have been made in this space since that time For anyone that's interested in what these may be. Take a peak at MANRS. https://www.manrs.org/ https://www.manrs.org/