Transcript CSO Coordination Meeting

CSO
Coordination Meeting
80th IETF, Prague
Agenda
• CSO IRTF Proposal (Young, 10 min)
• CSO and ALTO (Greg, 10 min)
• Carrier Perspective
– Network Capability (Javier, 5 minutes)
– Synchronized Recovery (Ning, 5 minutes)
– Application-Network Integration (Kohei, 5 minutes)
• Vendor Perspective (Paul, 5 minutes)
• Open Discussion (30 minutes)
• Wrap-up
80th IETF, Prague
Background
• This is NOT the same as previous conversation
in IETF.
• We are trying to identify clear research
problems associated with CSO in IRTF context.
– Is it possible to do network resource aware global
load balancing, improve responsiveness between
application and network and resiliency?
– What are the key building blocks that would
enable CSO?
CSO Context: Evolving relationship
between Data Centers and Networks
Server
Racks
Co-Lo
DWDM
AS
(L2)
Router DWDM
(CE)
Metro
ULH
DWDM
PE
P
ULH
ULH
P
PE
ULH
DWDM
Data Center
(a) Existing
Server
Racks
Co-Lo
OTP
AS Lambda/OTU/Ethernet OTN
MPLS TP
(CE)
PE
Data Center
ULH
ULH
ULH
PE
ULH
(b) New
• Significant changes in network access, e.g., from a drop-off of an optical metro ring (a wavelength
or two), to an end destination in a long haul DWDM system (many wavelengths, multiple fibers).
• Emerging applications require high bandwidth such as sporting events, live converts, 3D video
applications, remote medical surgery and so on.
• Network capability that provides flexible, dynamic and reliable transport needs to be considered
for mission-critical applications
80th IETF, Prague
Goals: Improve resource utilization, resilience,
and QoE for datacenter based network
applications
ACG
(GLB)
Application
(server)
CE 1
DC 1
User Request
R
Edge
-
CE 2
Core
Access
DC 2
Edge
SCG
Transport
Network
Core
-
Edge
Core
Core
Edge
Direct AccessR
(Corporate User)
DC 3
80th IETF, Prague
R
Direct Access
(Corporate User)
Key Research Problems under CSO
•
•
•
•
Joint application/network Resource optimization
(global load balancing)
Responsiveness to quickly changing demands
from/to application to/from network
Enhanced service resilience (via cooperative
recovery techniques between application and
network)
Quality of application experience (QoE)
enhancement (via better use of existing network
and application resources)
80th IETF, Prague
Tentative Deliverables
•
Develop baseline network/application architecture and
model – general enough to include most cases of interest.
– Define the various "trust or lack of" in the relationships between
various key players in both the application and network stratum.
– Define key interfaces and their functionality and relate these to
current standards and potential future standards.
•
Demonstrate feasibility of CSO concept via prototype and
experiments.
– Determine scalable mechanisms for application level support for
network recovery and network support for application recovery.
– Experiment network capabilities (e.g., IP, MPLS, GMPLS) in providing
support for dynamic application loads, scaling and QoE enhancement.
80th IETF, Prague