Boffins devise 'cyberweapon' to take down internet
Love BGP will tear us apart
University Boffins say they've devised a way to take down the internet by turning core parts of its routing protocol against itself.
The attack, which was presented last week at the Network and Distributed System Security Symposium in San Diego, California, attacks functionality in the BGP, or Border Gateway Protocol. The technology is designed to ensure that links between internet service providers and other large network operators never go down, by routing around paths that are no longer valid. Attackers would be required to have a botnet of about 250,000 infected machines to map major routes between ASes, or autonomous systems.
New Scientist, which describes the attack developed by University of Minnesota professor PhD candidate Max Schuchard here, goes on to say:
An attacker deploying the Schuchard cyberweapon would send traffic between computers in their botnet to build a map of the paths between them. Then they would identify a link common to many different paths and launch a ZMW attack to bring it down. Neighbouring routers would respond by sending out BGP updates to reroute traffic elsewhere. A short time later, the two sundered routers would reconnect and send out their own BGP updates, upon which attack traffic would start flowing in again, causing them to disconnect once more. This cycle would repeat, with the single breaking and reforming link sending out waves of BGP updates to every router on the internet. Eventually each router in the world would be receiving more updates than it could handle – after 20 minutes of attacking, a queue requiring 100 minutes of processing would have built up.
Schuchard went on to tell the publication it would take days for the collective internet to recover and fixes would only come about “by network operators actually talking to each other.” For the fix to work “each autonomous system would have to be taken down and rebooted to clear the BGP backlog.”
BGP was designed to keep email and other data moving over the internet even when paths between two or more stops are no longer working. Rather than scuttle the transmission, packets are simply rerouted on-the-fly over another route.
The attack, laid out in a paper titled Losing control of the internet: using the data plane to attack the control plane, is the latest to exploit weaknesses in BGP. In August 2008, researchers at the Defcon hacker conference outlined a technique for hijacking huge chunks of the internet that capitalized on the implicit trust placed in BGP routers.
That trust has also been exploited to divert internet communications – some of it from the US military – though China.
The theoretical assault described by Schuchard and his colleagues is essentially a denial-of-service attack that targets the internet's control plane. The researchers estimate with a 250,000-node botnet, “the median load on nearly half of the core routers increased by a factor of 20 or more.” ®
Regcast training : Hyper-V 3.0, VM high availability and disaster recovery
COMMENTS
seriously?
does it take in to account route flap dampening which all the major networks do as standard specifically to stop flapping routes causing a problem? if a link keeps flapping up and down then updates from it are suppressed and stop being propagated, this causes a slightly longer outage of the flapping link, but protects the rest of the internet from having to process the updates
route flapping isn't uncommon, if it caused widescale outages every time it happened then we'd have serious problems
A bit more than a router flap, I think...
> This sounds similar to our "Router flap' incident.
Well, if operators really had to isolate and reboot all the BGP4 speakers in the world, it would be a bit more of a news story than that, I think. $230/minute wouldn't come near to covering it.
However, I suspect there is more robustness in the real world than the paper assumes, and operators do talk to each other when things go funny. Admittedly, they usually talk by email, and they might have to revert to steam-powered telephones in this case.
losing money , or just not making it?
"And $230 a minute is micky mouse.We're a Midsize Enterprise and even we are talking £1m / hour. Happen to a bank and you're talking 100x that upwards."
100 million dollars per hour ? lost ?
sounds a little exagerrated in my completely uneducated guesstimate
I've never liked the capitalistic " were losing x per hour" no - your just not making x per hour at the moment. I remember it from that IBM ad where systum gon down
no wonder banks can lose our money so fast

IT infrastructure monitoring strategies
What you need to know about cloud backup
Agentless Backup is Not a Myth
Top 10 SIEM Implementer’s Checklist
Steps to Take Before Choosing a Business Continuity Partner