4 ms·

> necessity to have filters on both sides of an EBGP session. In this case it appears Verizon had little or no filters, and accepted most if not all BGP announc

> necessity to have filters on both sides of an EBGP session. In this case it appears Verizon had little or no filters, and accepted most if not all BGP announcements from Google which lead to widespread service disruptions. At the minimum Verizon should probably have a maximum-prefix limit on their side and perhaps some as-path filters which would have prevented the wide spread impact. Wow that's just stupid on Verizon's part given the magnitude of potential impacts.

akvadrako 9y ago
It's not that stupid; it's an intentional tradeoff. It's also quite common, especially with a customer like Google.
- blinkingled 9y ago
  Everybody (including Google as we find) makes mistakes and not having safeguards doesn't sound like a good trade-off to me. (What are they trading off here again?)
  - rnxrx 9y ago
    Operational complexity. For a carrier the size of Verizon to build an authoritative list of prefixes to advertise they would have to have a registration process for every single customer, peer, etc that was both reliable and up to the minute. That list would have to be compiled and driven on a near constant basis to many tens of thousands of routers and potentially hundreds of thousands of BGP sessions. It's not that this isn't doable, but it touches far more process in far more places than might be obvious and potentially requires additional capacity/capability on a lot of hardware. For a giant network this all means a great deal of time and money. Should they do it? Absolutely. Is it cheap/easy? Absolutely not. More cynically, does the potential cost and disruption of doing the right thing map favorably against their view of the risk of NOT doing it? Unknown.
    - jlgaddis 9y ago
      Fortunately, this problem was solved ages ago by central routing registries. I register my prefixes and have a single object that upstreams and peers can use to build their filters for BGP sessions with me. Why can't Google do the same? ("But they're soooo big" is not an acceptable answer.)
      - rnxrx 9y ago
        It was "solved" in some general sense but was never implemented widely or consistently. It was traditionally a really clunky and fragile system (e-mail forms) that ended up being really limited in its utility because it wasn't widely used. That said, the point isn't that this is a technically difficult problem (it isn't) but rather one of human scale: altering the behavior of tens of thousands of more-or-less autonomous networks spanning the globe is non-trivial.

akvadrako 9y ago

It's not that stupid; it's an intentional tradeoff. It's also quite common, especially with a customer like Google.

blinkingled 9y ago
Everybody (including Google as we find) makes mistakes and not having safeguards doesn't sound like a good trade-off to me. (What are they trading off here again?)
- rnxrx 9y ago
  Operational complexity. For a carrier the size of Verizon to build an authoritative list of prefixes to advertise they would have to have a registration process for every single customer, peer, etc that was both reliable and up to the minute. That list would have to be compiled and driven on a near constant basis to many tens of thousands of routers and potentially hundreds of thousands of BGP sessions. It's not that this isn't doable, but it touches far more process in far more places than might be obvious and potentially requires additional capacity/capability on a lot of hardware. For a giant network this all means a great deal of time and money. Should they do it? Absolutely. Is it cheap/easy? Absolutely not. More cynically, does the potential cost and disruption of doing the right thing map favorably against their view of the risk of NOT doing it? Unknown.
  - jlgaddis 9y ago
    Fortunately, this problem was solved ages ago by central routing registries. I register my prefixes and have a single object that upstreams and peers can use to build their filters for BGP sessions with me. Why can't Google do the same? ("But they're soooo big" is not an acceptable answer.)
    - rnxrx 9y ago
      It was "solved" in some general sense but was never implemented widely or consistently. It was traditionally a really clunky and fragile system (e-mail forms) that ended up being really limited in its utility because it wasn't widely used. That said, the point isn't that this is a technically difficult problem (it isn't) but rather one of human scale: altering the behavior of tens of thousands of more-or-less autonomous networks spanning the globe is non-trivial.