How critical is BGP MD5 at Internet Exchange Points? Would lack of support for MD5 authentication on route servers prevent some peers from multilaterally connecting? Do most exchange operators support it? Thanks! Jay
On Mar 9, 2012, at 17:24 , Jay Hanke wrote:
How critical is BGP MD5 at Internet Exchange Points? Would lack of support for MD5 authentication on route servers prevent some peers from multilaterally connecting? Do most exchange operators support it?
<http://mailman.nanog.org/pipermail/nanog/2012-January/thread.html> Search for MD5. Most IXP route servers support it, few require it. So even if you do it on your side, doesn't mean someone else did it on their side. I've never seen anyone refuse to connect to an IXP route server that did not, but that doesn't mean it hasn't happened. -- TTFN, patrick
On 9 Mar 2012, at 22:24, Jay Hanke wrote:
How critical is BGP MD5 at Internet Exchange Points? Would lack of support for MD5 authentication on route servers prevent some peers from multilaterally connecting? Do most exchange operators support it?
At LONAP in London, the route-servers do not support TCP MD5 authentication for BGP. i don't think that this policy has led to anyone refusing to connect (about 80 of the 110 or so peers connected to the exchange use the Multilateral service - it is optional to connect to the MLP). We have no plans to enable TCP MD5 on this service. Because TCP MD5 packets touch a router's CPU, using MD5 introduces a new attack vector - see nanogii passim (e.g. http://www.nanog.org/meetings/nanog39/presentations/Scholl.pdf). Don't do it. :-) Andy
Andy Davidson <andy@nosignal.org> writes:
Because TCP MD5 packets touch a router's CPU, using MD5 introduces a new attack vector - see nanogii passim (e.g. http://www.nanog.org/meetings/nanog39/presentations/Scholl.pdf). Don't do it. :-)
Tom's slide deck is often misinterpreted - the salient parts are on pages 13 and 16. The big problem is that random packets can hit the control plane - AT ALL. One can kill it just as easily with a synflood on some other tcp port (perhaps ssh or even one that it isn't listening on?). Hopefully your modern exchange point router has some sort of control plane policing. Indeed, hopefully your backbone routers have some sort of control plane policing mechanism and you have turned it on and subjected your policy to some scrutiny. Blowing up un-password-protected BGP sessions by spoofing has not turned out to be a big problem in recent years. It probably helps that the dangers of highly predictable ISNs have become well-known (and hopefully acted upon by router vendors but history has shown that making blanket statements about that sort of thing on NANOG is unwise). A read of http://tools.ietf.org/html/rfc6528 may prove instructional. Turning on or off md5 makes minimal difference in CPU loading either during an attack or not, but it is another thing to get wrong - operational complexity without significant reward. I agree with Andy's conclusion. Don't do it unless whoever you're peering with demands it. It's not worth the complexity to set it up in the first place, and it's not worth your time to argue against it if someone is quite convinced that enabling md5 on your bgp session will save the world. -r
On 10/03/2012 11:24, Robert E. Seastrom wrote:
Hopefully your modern exchange point router has some sort of control plane policing.
My gut feeling is that lots don't. The behaviour of various operating systems regarding MD5 processing is interesting. *BSD (and I assume consequently junos) checks ttl and sequence numbers before checking md5. Linux and IOS do md5 first, and I just wonder about the wisdom of this approach due to the slightly higher computational overhead of calculating the hash. In general, I'm slightly in favour of md5 at ixps, not because of session security, but when exchange participants leave an ixp, lots of people don't bother to remove the bgp sessions. If as a newcomer to the IXP you get a re-used ip address, without md5 it can sometimes be possible to do Interesting and Bad Things with old sessions from other ixp participants. FWIW, for the INEX route server system we: - use bsd - implement packet filtering to accept tcp/bgp only from the ixp subnet - generally use md5 for ipv4 sessions - generally don't use md5 for ipv6 sessions for historical reasons This works for us.
I agree with Andy's conclusion. Don't do it unless whoever you're peering with demands it. It's not worth the complexity to set it up in the first place, and it's not worth your time to argue against it if someone is quite convinced that enabling md5 on your bgp session will save the world.
yep, agreed. Doesn't make that much difference in real life so don't lose sleep about it. The only real difference it makes is that it can help shut up "security" audit people (the tick-box compliance variety) from their ivory tower whining. Nick
participants (5)
-
Andy Davidson
-
Jay Hanke
-
Nick Hilliard
-
Patrick W. Gilmore
-
Robert E. Seastrom