Is IP going to take over the world (of communications)? Pablo Molinero-Fernandez Stanford University Nick McKeown Stanford University Hui Zhang Turin Networks, Carnegie Mellon University Text from Alan Mislove, Ansley Post
Background • The Internet is one of the most successful communications platforms because of two characteristics • Reachability • Heterogeneity • Almost all Internet traffic is over Internet Protocol (IP) • First created in 1970, called Mark I
Background (cont) • Internet’s success has lead to the assumption that it will become the only communication platform • Phone Networks would be replaced with voice-over-IP systems; • Skype (100 million users in just 3 years) • Vonage (100k users in just 3 years), • Movies and Television will be distributed using Internet • YouTube, VideoFurnace, Google Videos • Similarly, it has also lead to the assumption that packet-switching (IP) routers will become the only switching device.
Motivation • Despite its strengths, IP is not necessarily the best solution • Goal of our paper: • Counter the assumption that IP will “take over the world (of communications)” • Comparing packet switching to other switching devices • Shatter false assumptions that lead to this belief
IP Misbeliefs • Widely held assumptions that will be questioned • Current dominance of IP over other communications • IP’s efficiency • IP’s robustness • IP’s simplicity • Real-time applications using IP
Dominance of IP • It is falsely believed that IP currently dominates global communication • ISP’s have revenues of $13B • Compared to other communications’ revenues totaling $344B • Internet only reaches 59% of US • phone 94% & TV 98% • IP market is smaller in the data & telephony infrastructure • IP routers market size: $4.1B • circuit-based router market size: $32.5B
Efficiency of IP • IP makes efficient use of “scarce” bandwidth • But is it really scarce? • Average Internet link utilization is 3%-20% • Ethernet networks utilization is much lower at ~1% • Long-distance phone line utilization is 33% • Network’s over provision allows for a low packet delay
Efficiency of IP (cont) • Why? (Over Provision) • Internet traffic is asymmetric and bursty • Difficult to predict traffic growth on a link • Economical to add large increments of capacity • However, there are “less talked-about” reasons • Under congestion, IP performs badly • Control traffic transmitted in-band • Results in black holes, traffic loops, etc… • Much easier to keep utilization low
Efficiency of IP (cont) • In practice, user experiences the same delay in packet-switched or circuit-switched network • Average user’s work (65%) is request-response • File sharing • For most types of workloads, circuit-switching provides same user response time
Robustness of IP • Internet was designed to withstand catastrophic event. • Median Internet downtime is 471 minutes/year • Median phone downtime is 5 minutes/year • BGP convergence is slow (3-15 minutes) • SONET/SDH switches to a backup path in 50ms • Nothing inherently unreliable about circuit-switching
Simplicity of IP • Beginning principle is that complexity should be at the endpoints • Increasingly, IP routers have become sophisticated • Multicast • Quality of Service • VPN • Configuring IP routers can be very difficult • Single misconfigured IP router can cause instability for a large portion of the network
Simplicity of IP (cont.) • IP routers have about 8 million • Circuit-switched routers have 3 million lines of code • IP routers have 300 million gates, 1 CPU, 300 MB of buffer space • Circuit routers have 25% of the gates and no CPU • Circuit-switched routers sell for 9% - 50% the price • IP routers do not lend themselves to implementing optics • Circuit switching is compatible with optical technology
Real-Time Support in IP • Widely held assumption that IP will support real-time applications • This assumption relies on over provisioning of the network • Or Quality-of-service in the network that has yet to be implemented • Even after 10 years of research, no infrastructure has been created (using IP routers) to replace circuit switching.
What if we started over? • Hybrid solution would be most appropriate • Uses packet switching at the edges • Circuit-switching at the core and with applications with QoS demands • Tightly integrate these two parts
Conclusion • IP does some things well, but not everything • Good for scarce-bandwidth situations • Wireless, undersea cables, satellite links • Inappropriate for real-time applications • Voice traffic, telephony • If we redesigned the Internet, not all routers would be packet-switching • Core routers and real-time application data would be circuit-switched
Questions? • Mike O’Dell, former Senior VP, UUNet: • “[to have a voice-over-IP network service one has to] create the most expensive data service to run an application for which people are willing to pay less money everyday, […] and for which telephony already provides a better solution with a marginal cost of almost zero”