Stanford Clean Slate Internet Design Program

1
Stanford Clean Slate Internet Design Program

• Reinventing the Internet

Guru Parulkar, Executive Director, Clean Slate
Internet Design Funded by Cisco, Deutsche
Telekom, DoCoMo, Ericsson, LightSpeed, MDV, NEC,
NSF, Xilinx
2
What is Common Among These Companies?
3
BIG CHANGES ON THE HORIZON
4
Revolution in Mobile Computing Millions g
Billions
Power-limitation of handheld a computation will
move to the cloud Need to back up and refresh our
lost data a data will move to the cloud
5
Computing and Storage Moving Into the Cloud
6
Cyber-Physical World
New Machines
7
Network Centric Critical Infrastructures
Essential Utilities
Internet Architectureis NOT robust enough to
support these
Transportation
Telecommunications Banking Finance
8
Persistent Problems Making Internet
Infrastructure Worse
in the thirty-odd years since its invention,
new uses and abuses, , are pushing the Internet
into realms that its original design neither
anticipated nor easily accommodates. Freezing
forevermore the current architecture would be bad
enough, but in fact the situation is
deteriorating. Overcoming Barriers to
Disruptive Innovation in Networking, NSF Workshop
Report, 05.
9
BIG CHANGES REPRESENT BIG OPPORTUNITIES
10
Big Changes Represent Big Opportunities

• Opportunities for
• Research groups to shape future Internet
• Startups to create new product categories
• Incumbents to get into new markets and grow
• Newcomers to leapfrog
• Not the time to sit on the sidelines

11
Stanford Clean Slate Program

• To reinvent Internet infrastructure and services
  
• by creating key platforms for innovations and
  deploying and making them available to research
  and user communities
• with emphasis on mobile computing

12
Stanford Team
Education
P. Kim
Applications
L. Guibas
B. Girod
S. Klemmer
HCI
D. Boneh
Security
J. Mitchell
Languages
J. Ousterhout
M. Lam
Distributed Systems
D. Mazieres
V. Koltun
P. Levis
OS
M. Rosenblum
Architecture
K. Kozyrakis
Economics
R. Johari
G. Parulkar
N. McKeown
Networking
F. Tobagi
B. Prabhakar L. Kazovsky
A. Goldsmith
Radio
A. Paulraj
13
Example Projects
Architectural Blueprint of Future Internet?
Prog Open Mobile Internet(POMI) 2020
??
14
Revolution in Mobile Computing Millions g
Billions
Power-limitation of handheld a computation will
move to the cloud Need to back up and refresh our
lost data a data will move to the cloud
15
Today
Vision

• Problem with the network.
• 3G Cellular networks a IP
• IP Bad for mobility, security, management
• Need a network that continually evolves

• When theyve got our data, theyve got us!

• Surrounded by capacity we cant use
• Inefficient Costs more, poorer quality
• We need an alternative

• Big-brother portals luring us to their
  repository
• We have to provide an alternative
• Healthcare, Financial May never take off

Where we will go otherwise
16
Stanford Clean Slate Program

• To reinvent Internet infrastructure and services
  
• by creating key platforms for innovations and
  deploying and making them available to research
  and user communities
• with emphasis on mobile computing

17
The Big Picture
Applications PocketSchool, Virtual Worlds,
Augmented Reality
WEB/Computing Substrate Network of VMs, Mobile VMs
Economics
Data Substrate PRPL Virtual Data System
Network Substrate OpenFlow
Radio technologyMulti-Gb/s, 99 coverage
18
Mobile Network Infrastructure Today

• Many cellular networks visible (5-7 common), many
  wifi networks visible (10-15 common).
• But not practically available to me closed
  infrastructures.
• Seamless mobility impossible
• Why cant I use of all the infrastructure around
  me?

WiFi
cellular
19
Network Substrate Goals

• Access to all infrastructure
• Continued connectivity and seamless mobility as I
  move
• Maximize user choice
• Radio
• Handoff
• Allow innovation
• Handoff mechanisms
• AAA, billing,

20
OpenFlow Basics
PC
OpenFlow Switch
OpenFlow Protocol
SSL
Controller
Secure Channel
sw

• Add/delete flow entries
• Encapsulated packets
• Controller discovery

Flow Table
hw
21
Example Network Services

• Static VLANs
• New routing protocol unicast, multicast,
  multipath,
• Network access control
• Home network manager
• Mobility manager
• Energy manager
• Packet processor (in controller)
• IPvX
• Network measurement and visualization
• These other services gt a very different net
  substrate

22
SIGCOMM 2008 Demo
23
Virtualized OpenFlow Substrate
Hypervisor Policy Control
OpenFlow Switch
OpenFlow Protocol
OpenFlow Switch
OpenFlow Switch
24
OpenFlow and Mobility

• Lots of interesting questions
• Management of flows
• Control of switches
• Access control of users and devices
• Tracking user location and motion

• Lots of radio networksWiFi, WiMax, LTE,
• Dumb access points
• User choice

25
Deployment on Stanford campus

• 100 of WiFi APs in 4 buildings outdoor
  locations
• A few Mobile WiMAX femto-cellbase stations
• Deployed in this autumn
• All are OpenFlow enabled connected by OpenFlow
  switches
• Plan to have a project class in this
  autumn/winter quarter

WiFi AP (two radios/box)
We are ready for innovation in our network!
Mobile WiMAX AP
26
Broader Impact OpenFlow Network Substrate

• Eight switch vendors enabling this capability
• Cisco, HP, NEC, Juniper, and others
• We are starting to demonstrate the key
  capabilities
• ACM SIGCOMM08
• GENI Engineering Conference
• Supercomputing
• We are deploying
• on our campus two buildings at Stanford
  (HP/Cisco)
• on other campuses in US and Japan with NSF
  support
• in national nets US (Internet2, NLR), Japan
  (JGN2plus), Europe,
• And enable researchers and network operators to
  innovate on topHope OpenFlow takes off -- on a
  path of no return

27
The Big Picture
Applications PocketSchool, Virtual Worlds,
Augmented Reality
WEB/Computing Substrate Network of VMs, Mobile VMs
Economics
Data Substrate PRPL Virtual Data System
Network Substrate OpenFlow
Radio technologyMulti-Gb/s, 99 coverage
28
Do you know where your data are?

• Our data
• moving into the cloud
• owned by applications and users losing control
• difficult to share among applications
• leakage a serious problems
• Trends to accelerate unless checked

29
PRPL PRivate-PubLic Data Index
Old New Data Apps/Services
PRPL

• A unified view of data
• Separate data ownership, storage, applications
• Secure, fine-grain sharing
• Device-independence caching
• Interactive data navigation with semantic-web
  queries

30
The Big Picture
Applications PocketSchool, Virtual Worlds,
Augmented Reality
WEB/Computing Substrate Network of VMs, Mobile VMs
Economics
Data Substrate PRPL Virtual Data System
Network Substrate OpenFlow
Radio technologyMulti-Gb/s, 99 coverage
31
Existing WEB/Computing Substrate

• Incremental progress document sharing to default
  for all web apps
• Not designed for all its current and future
  applications
• Scale in terms of type of content, users, data
  repositories,
• Modularity, reusability, interactivity

32
Clean Slate WEB Framework Fiz

• Enable developer to create applications in
  days/weeks with
• reusable modular components
• high interactivity
• scalability to millions of users on diverse
  devices
• dynamic and rich content
• shared computing storage in the infrastructure

33
Fiz Architecture
URLs
Interactor
Interactor
Sections
Tree
Tabs
DataRequests
Front End
Back End
DataManager
DataManager
DataManager
Remote Feed
EnterpriseApplication
SQL Database
34
Execution Environment Centralized to Distributed
S
S

• Lots of interesting questions
• How to decompose a web app?
• What to cache and when to cache?
• What type of data consistency?
• How to interface to PRPL?

S
S
S
S

• High interactivity, low latency, mobile devices,
  and multiple data repositories require
  distributed implementation

35
Path to Broader Impact Fiz

• Initial version already taking shape
• Based on Java servlets Tomcat
• Make open-source releases
• Starting in summer 09?
• Create a Fiz community Web site
• Using Fiz itself
• Encourage creation of additional components
• Biggest challenge keeping it simple
• Iterate it may take several tries to get this
  right
• Major revisions?
• Start again from scratch?

36
Bringing Substrates Together
37
Content
38
(No Transcript)
39
Data Substrate PRPL Virtual Data System
Computation Substrate Network of VMs, Mobile VMs
Network Substrate OpenFlow
Radio technologyMulti-Gb/s, 99 coverage
Energy aware OS
Content
UI
Client
OS
40
Summary

• Big changes on the horizon
• Opportunity to rethink the Internet
  infrastructure
• Stanfords Clean Slate Program
• Reinvent the Internet by creating platforms for
  innovations
• WEB/Computing substrate Fiz and network of VMs
• Enable scalable, highly interactive, rich media
  applications
• Data substrate PRPL Platform
• Separate data from applications in cloud give
  control of data to owners
• Allow any application to use any data under the
  control of its owner
• Networking substrate OpenFlow Platform
• Enable users to create their own network services
  
• Network services access control, routing,
  mobility management,
• Handheld software OS, browser, UI platforms