class: center, middle, contrast, narrow .displayinlineblock.width50.trans80[] ## a future Internet architecture that you can use today   <p>Giacomo Giuliari, Mateusz Kowalski,<br /> Juan A. Garcia Pardo, Kamila Součková</p> VIScon, 12 Oct 2019 --- # SCION future Internet architecture -- → changing the Internet is really hard, so… Why? -- .width100[] ??? Cloudflare, Google, Amazon, Reddit... --- .width100[] .pushuplots.center[ BGP route leak from the summer only one of many examples another one: in June a lot of European traffic went through China ] --- # What is wrong with the Internet? originally a research network * access was not widespread, everyone trusted everyone * if it broke for a day, nobody cared * end hosts were slow and stupid -- Today​: * many entities, they don’t trust each other * if it breaks for a few hours, everybody cares * end hosts are fast and smart, performance matters ??? * everyone has access and everyone has data on it --- # What is wrong with the Internet? In the future: * automation ⇒ internet will really become critical infrastructure * drones, self-driving cars, remote robots... * more and more data online ⇒ trust will be even more of an issue * geofencing (requiring data to stay within the country) * attacks will get harder to stop * more data transfers and faster hosts => network will be the bottleneck --- .floatright.width35[] # How does the Internet work today? you write destination IP address on packet and send it off .floatleft.width30[] * no control over **how** it gets there * no possibility of using multiple paths ??? there was bgp stuff here, and the pakistan thing, but maybe I won't even mention it? --- class: bottom, center **SCION** is designed for **.hl[route control]**, **.hl[failure isolation]**, and **.hl[explicit trust information]** for end-to-end communication   .floatleft.width60[] --- class: expanded1 # Isolation Domains (ISDs) .floatright.width60[] * e.g. ISD ≈ country * trust is per ISD * routing, certificates, DNS, etc. * routing within ISD independent * malicious actors outside of your ISD cannot do anything --- exclude: true class: expanded1 .floatright.width50[] # Path discovery * core ASes create *path construction beacons* (PCBs) & flood * hierarchical: * inter-ISD: among core ASes * intra-ISD: only downstream * AS receiving beacon adds itself & forwards it * ⇒ PCB contains known good path to core * adds its signature ⇒ PCBs can be verified --- exclude: true # Path discovery * AS chooses how to extend / what to forward ⇒ enables routing policies * available *path segments* registered to *path servers* * end hosts request from path servers * segments combined into end-to-end paths: --- # Path awareness * you choose the path, write it on the packet * routers just follow the instructions ⇒ more efficient * efficient cryptographic checks to ensure that you are allowed to use this path * multipath Just Works™ .center[ .displayinlineblock.width25[] .displayinlineblock.width25[] .displayinlineblock.width25[] ] ??? allowed to use this path: for DDoS defence and policy compliance and traffic engineering .center[ .displayinlineblock.width30[] .displayinlineblock.width30[] .displayinlineblock.width30[] ] --- exclude: true # Path discovery every PCB contains *hop fields* for path construction: * ingress, egress interface * timestamp + expiration time * chained cryptographic MACs with AS-specific keys ⇒ paths cannot be "invented" .width90[] --- exclude: true class: expanded # Packet forwarding .floatright.width60[] * path in packet header, no routing tables * every border router processes only its own *hop field* * checks to ensure path authenticity --- exclude: true class: circlist, expanded1 .floatright.width35[] # Secure by design 1. provable security: protocol + code 1. heterogeneous trust model: ISDs 1. path-aware networking: client chooses path 1. new routing + forwarding architecture: beaconing; packet-carried forwarding state 1. built-in DDoS defence mechanisms: hidden paths; source authentication; bandwidth reservation --- exclude: true # Incremental deployment * SCION-IP gateway (SIG) enables legacy hosts to transparently use SCION * SCION over IP enables "mixed" networks .centerblock.width95[] --- exclude: true # Deployment status * incremental deployment possible * backward compatibility with IP * possibility to reuse infrastructure * research (SCIONLab): * ISPs: Swisscom, SWITCH, KDDI, GEANT, DFN * Deployed 50 ASes worldwide * Global interest, e.g., ESA * commercial (Anapaya): * ISPs: Deutsche Telekom, Swisscom, SWITCH, Init7 * Bank deployment: 4 major Swiss banks, some in production use * Swiss government has SCION in production use --- # SCIONLab research network .width100[] .center[50 points of presence + more than a hundred user ASes] .center[world-wide collaboration: universities, ISPs, others] --- # SCION commercial development * Anapaya Systems founded in 2017 as ETH Zurich spinoff * current customers: multiple banks, ISPs, and more * tight collaboration with ETH .center[ .width40[] https://www.anapaya.net ] --- class: expanded1 # Say hi! .floatright.width20.pushup[] * https://netsec.ethz.ch Network Security Group at ETH * https://www.scion-architecture.net book, papers, videos, tutorials * https://www.scionlab.org * SCIONLab * https://github.com/scionproto/scion source code --- # SCION: future Internet that you can use today .centerblock.pure-table.small[ | | IP+BGP | SCION | | -------------------------------- |:-------------:|:-------------:| | path control + geofencing | throw and pray | you select path | | trust model | all or nothing | explicit trust | | scalability of network | routing tables full | ✓ | | availability / failover behaviour | slow convergence | no convergence | | path selection + multipath | ✗ | ✓ | | used by applications | ✓ | ✗ | ] --- # SCION: future Internet that you can use today .centerblock.pure-table.small[ | | IP+BGP | SCION | | -------------------------------- |:-------------:|:-------------:| | path control + geofencing | throw and pray | you select path | | trust model | all or nothing | explicit trust | | scalability of network | routing tables full | ✓ | | availability / failover behaviour | slow convergence | no convergence | | .hl[path selection + multipath] | ✗ | ✓ | | .hl[used by applications] | ✓ | ✗ | ] .center.hl.large[**Now it's your turn!**]