Impact of

1
Impact of application empowered networks
Question 3

• The semi-conductor revolution reduced CAPEX and
  OPEX costs for main frame computer
• But its biggest impact was not on main frame
  computer but allowing the creation of
  mini-computers, PCs, mobile phones, etc
• In the 1960s computing was a service like
  telecom is today
• Large mainframes leased by companies like IBM
• It was inconceivable that an individual could own
  a computer
• The same revolution may be about to happen to
  optical networking as with semi-conductor because
  of low cost optics
• Yes it will make networks cheaper for carriers to
  operate
• But the real revolution is in allowing us to
  think of new network models and architectures e.g
• Application empowered networks, customer
  controlled networks, overlay networks, etc

2
CAnet 4 is NOT an optical network
Question 3

• CAnet 4 is made up of many parallel application
  empowered or customer empowered specific networks
  eg
• Computer back planes (Westgrid)
• High energy physics network
• It extends the Internet 2 architecture of
  GigaPOPs connecting a small number of RE
  institutions to a much finer scale with many
  parallel application empowered Internet 2 like
  networks connecting individual researchers and/or
  applications
• With added feature that the application or user
  can dynamically manage their own IP network
  topology
• Application empowered networks peer with each
  other at GigaPOPs and at optical switches which
  provides for greater reliability
• The CAnet 4 wavelengths and switches are
  partitioned such that application empowered
  networks can control their own partition and
  incorporate alarms, topology and discovery into
  their IP network
• New ITU draft standard Y.1312 - Layer 1 VPNs

3
Drivers for application empowered networks -1
Question 1

• Distributed back planes between HPC Grid centers
• Westgrid 1 GbE moving to 10 GbE
• SHARCnet 10 GbE
• Distributed Single Mount file systems Yotta,
  Yotta - SGI
• Needs very consistent performance and throughput
  to truly act as a back plane
• Frequent topology changes to meet needs of
  specific applications
• Canada ATLAS 980 Gbytes FCAL data once a month
  from CERN to Carleton U, UoAlberta, UoArizona,
  etc
• Will significantly increase to Terabytes when
  production runs start
• Would take over 80 days on IP RE network

4
Drivers for application empowered networks - 2

• CERN Low level trigger data to UoAlberta with
  GARDEN
• Initially streaming data rates 1 Gbps moving to
  10Gbps later in the year
• Canadian virtual observatory
• .5 Tbyte per day to UoToronto and UoHawaii
• 250 Mbps continuous streaming from CCD devices
• Neptune Canada (and US?) under sea laboratory
  multiple HDTV cameras and sensors on sea floor
• Canada Light Source Synchrotron remote
  streaming of data acquisition to UoAlberta
• 2 to 5 Gbps continuously
• Canadian remote Nano and micro electronics
  laboratories

5
A VPN alternative to GMPLSLayer 1 VPNs
Question 3

• Allows customer to create customer owned and
  managed networks with resource heterogeneity
• Integration of wavelengths and dark fiber from
  different carriers
• Customers can manage their own restoral and
  protection schemes
• Customer can create daughter VPNs and offer to
  other users
• Customer can autonomously connect VPNs with other
  third parties

6
Enables new network architectures
Question 3

• Eliminate expensive high end routers and replace
  them with partial mesh of lightpaths between edge
  routers and servers
• But circuits are NOT intended to replace packet
  networks
• Extend the Internet end to end principle to the
  topology layer
• The success of the Internet is largely
  attributable to the classic e2e principle where
  control is at the edge
• Users can now control topology as well as
  applications
• Allowed development of exciting new applications
  or services
• Many exciting new overlay networks
• Knowledgeplane
• Oceanstore, Chord
• PlanetLab
• Application empowered networks allow overlay
  network to optimized underlay topology

7
The GigaPOP concept
Question 3
University
University
Commodity Internet
GigaPOP
GigaPOP
University
University
vBNS
University
University
8
CAnet 4 Internet 3?
Question 3 Question 4
University
Dept
UltraLight
CERN
Commodity Internet
University
University
vBNS
GigaPOP
GigaPOP
University
University
eVBLI
9
Hypothetical UltraLight Config
Question 3
CERN
User controlled topology
UltraLight
GigaPOP
GigaPOP
Caltech
10
Govt Funding
Question 4

• Governments are increasingly unlikely to fund
  general purpose IP research networks
• General purpose IP research networks are large
  vectors for DOS attacks and music/video file
  sharing
• They want to see a stronger coupling between the
  network and specific research activity
• Many new research applications have data flows
  that dwarf the general purpose IP RE network
• Application empowered networks will be discipline
  or application specific e.g. Ultralight, e-VBLI