Transcript Revision - May 5, 2009

Revision to DOE proposal
Resource Optimization in Hybrid Core
Networks with 100G Links
Original submission: April 30, 2009
Date: May 4, 2009
PI: Malathi Veeraraghavan
University of Virginia
1
Revisions
(serves as Outline)
•
•
•
•
•
•
•
•
•
•
•
Rev. 1: Generalization from 100G and lambdas
Rev. 2: Bidirectional resource optimization
Rev. 3: Hybrid node and hybrid network arch.
Rev. 4: Creation of bypass circuits
Rev. 5: Control-plane interaction with nodes
Rev. 6: Triggering of circuit setup and release
Rev. 7: Extensions of OSCARS
Rev. 8: DOE-provided testbed
Rev. 9: Areas of focus
Rev. 10: Policy issues
Rev. 11: Modified deliverables
2
Rev. 1:
Generalization from 100G and lambdas
• Original proposed work:
– 100G interfaces and optical WDM
– High-capacity switches/routers
• Modification:
– Improve understanding of hybrid network
operation at arbitrary rates, e.g., even at
10Gb/s or lower rates
– Circuits are generic and not necessarily only
wavelengths (lambdas); they can be sub-Gbps
SONET circuits, MPLS LSPs or carrier-grade
Ethernet virtual circuits
3
Rev. 2:
Bidirectional resource optimization
• Original proposed work:
– IP-routed traffic  Dynamic circuits
– Set up or release dynamic circuits in response
to surges or drops in IP-routed traffic
• Modification:
– Add opposite direction:
• Dynamic circuits  IP-routed traffic
– If a circuit setup is blocked due to a lack of
resources, the flow is sent on the IP-routed
path or MPLS LSP if the user request allows
this option, i.e., user is willing to accept sub-par
4
performance instead of being blocked
Rev. 3:
Hybrid node structure
• Original proposed work:
– Fig. 2b shows an IP router and a circuit
switch at each PoP with a pooled set of
lambdas interconnecting the circuit
switches at the PoPs
• Modification:
– A hybrid node could be one entity with
support for IP Layer-3 packet
forwarding and dynamic circuit switching
(layers 2.5/2/1); see the next slide
5
Rev. 3: Hybrid node architecture
Unfolded view: switch capabilities
Ethernet
control-plane
port
Node controller
Signaling
(provisioning)
protocol
SNMP
MIB+agents
Routing protocol
Hybrid node
D: Demultiplexer
M: Multiplexer
Administrative
interface (CLI, TL1)
Input
interfaces
1
2
3
Q-1
Q
Data plane
Layer-3
IP router
D
D
Layer-2/2.5
switch
D
D
D
Output
interfaces
M
M
M
Layer-1
switch
M
M
1
2
3
Q-1
Q
6
Rev. 3: Hybrid network architecture
(systems in blue to be implemented in this project)
(view with animation)
“put the traffic where the bandwidth is”
REF
Hybrid Traffic engineering (TE) system
Modify routing metrics and/or
write routing table entries
Hybrid Network Engineering (NE) system
Traffic monitoring/
characterization system
K circuits: IP-routed partition
N-K: Dynamic-circuit partition
Hybrid
Node
“put the bandwidth where the traffic is”
Obtain data
Request dynamic circuit
setup/release
DOE-implemented control-plane
software systems
Hybrid
Node
Hybrid
Node
Shared single core pool
of N fibers
Hybrid
Node
Hybrid
Node
REF: Report of US/EU Workshop on Key Issues and Grand Challenges in Optical Networking.
Available: http://networks.cs.ucdavis.edu/mukherje/US-EU-wksp-June05-Final-Report.pdf
7
Rev. 3: Hybrid network architecture
(systems in blue to be implemented in this project)
Hybrid Traffic engineering (TE) system
Hybrid Network Engineering (NE) system
Traffic monitoring/
characterization system
DOE-implemented control-plane
software systems
Traffic monitoring/characterization system
K circuits: IP-routed partition
• Reads parameters necessary only for TE/NE applications
N-K: Dynamic-circuit partition
Hybrid
• Characterizes traffic matrix
Node
Hybrid
• Not itself a general-purpose
monitoring system such as PerfSONAR
but could interface with such systems to obtain data
Node
Hybrid
Node
Shared single core pool
of N fibers
Hybrid
Node
Hybrid
Node
Report of US/EU Workshop on Key Issues and Grand Challenges in Optical Networking.
Available: http://networks.cs.ucdavis.edu/mukherje/US-EU-wksp-June05-Final-Report.pdf
8
Rev. 3: Hybrid network architecture
(systems in blue to be implemented in this project)
Hybrid Traffic engineering (TE) system
Hybrid Network Engineering (NE) system
Traffic monitoring/
characterization system
DOE-implemented control-plane
software systems
Hybrid Traffic Engineering (TE) system
K circuits: IP-routed partition
• Obtains data from Traffic monitoring/characterization system
N-K: Dynamic-circuit partition
Hybrid
• Computes optimal routes for load balancing
Hybrid
• Issues CLI commands toNode
hybrid nodes to modify routing metrics
and/or write routing table entries
Node
Hybrid
Node
Shared single core pool
of N fibers
Hybrid
Node
Hybrid
Node
Report of US/EU Workshop on Key Issues and Grand Challenges in Optical Networking.
Available: http://networks.cs.ucdavis.edu/mukherje/US-EU-wksp-June05-Final-Report.pdf
9
Rev. 3: Hybrid network architecture
(systems in blue to be implemented in this project)
Hybrid Traffic engineering (TE) system
Hybrid Network Engineering (NE) system
Traffic monitoring/
characterization system
DOE-implemented control-plane
software systems
Hybrid Network Engineering (NE) system
K circuits: IP-routed partition
• Obtains
monitoring/characterization system and
N-K: Dynamic-circuit
partitiondata from Traffic
Hybrid
DOE-implemented control-plane
Node software
Hybrid
• Determines if thresholds are crossed to trigger setup/release of
Node
dynamic circuits
Hybrid • If triggered, sends request for dynamic circuit setup/release to
DOE-implemented
software
Sharedcontrol-plane
single core
pool
Node
• Commands Hybrid TE system to make routing table updates
of N fibers
Hybrid
Node
Hybrid
Node
Report of US/EU Workshop on Key Issues and Grand Challenges in Optical Networking.
Available: http://networks.cs.ucdavis.edu/mukherje/US-EU-wksp-June05-Final-Report.pdf
10
Rev. 4:
Creation of bypass circuits
• Traffic-monitoring software monitors IP
traffic as well as dynamic circuit traffic
– If certain criteria are reached
(“thresholds”), then signal network and/or
traffic engineering to
• move IP-routed traffic going to a specific egress
node onto newly created circuits avoiding
intermediate routers
• move traffic from existing circuits onto the IProuted network
11
Rev. 5:
Control-plane interactions
Hybrid network
Traffic engineering system
Network engineering system
Traffic monitoring &
characterization system
Obtain topology and Request dynamic circuit
TE-database/PCE data
setup/release
DOE-implemented control plane
software systems
OSCARS InterDomain Controller
Domain
Controllers
Hybrid node
Hybrid node
•
Hybrid node
Collect data from domain controllers on capacity availability on various links to
enable path computation for new circuits triggered as a result of traffic
monitoring input
12
Rev. 5:
Control-plane messaging
• Control-plane messages are carried over the IProuted network
• Switch control cards have Ethernet control-plane
ports for connection into the IP-routed network
• For security, ns5 or equivalent IPsec devices
should be deployed on these control ports
13
Rev. 6:
Triggering of circuit setup and release
• Determine the criteria and thresholds for
triggering dynamic circuit setup in networkengineering module (using input from traffic
monitoring + DOE-implemented control-plane
software)
• Determine the criteria and threshold for
triggering circuit release
– cannot wait for traffic on a 10Gb/s circuit to fall to 0
before triggering release
– what should the threshold be?
• Determine the policies that govern the
implementation of thresholds
14
Rev. 7:
Extension of OSCARS to Layer 1-2 networks
• OSCARS scheduler was developed for
advance reservation of dynamic MPLS
(layer 2.5) LSPs
• Extension to Layer 1-2 networks
– Layer-2 (Carrier-Ethernet and SONET)
and Layer-1 (optical WDM and fiber)
– Develop algorithms suitable for
reservations and provisioning of nested
and concatenated circuits
15
Rev. 8:
DOE-provided testbed
• Demonstrate operation of hybrid networks
with resource optimizing hybrid TE and
hybrid NE software on DOE-provided
testbed with 3 to 4 nodes.
– Hybrid nodes (provided to us):
• IP router (layer 3) + MPLS switch (layer 2.5)
• Layer 1/layer 2 circuit switch
• DOE-implemented control-plane software
– Implemented by us:
• Hybrid TE, Hybrid NE and and traffic
monitoring/characterization software
16
Rev. 9:
Areas of focus
• Theoretical framework
– Use of traffic engineering and network
engineering in hybrid networks for
resource optimization
• Prototype demonstrations on DOEprovided testbed
17
Rev. 10:
Policy issues
• Should users be blocked if resources are
not available for circuits?
• Can users indicate option to avoid being
blocked but rather to use IP-routed/MPLS
LSP path if Layer 1 or Layer 2 circuit
network has no resources?
• Should decision be made at the edge or
inside the network?
• Is the decision driven by TE process or by
users or both?
18
Rev. 11:
Modified annual deliverables
(quarterly on next three slides)
• Year 1: Architecture and Analysis
• Year 2: Algorithm design and
software implementation
• Year 3: Prototyping on DOE-provided
testbed
19
Year 1 quarterly deliverables
• Year 1: Architecture and Analysis
– Q1: Design architectural framework for
hybrid networks with automated traffic
and network engineering
– Q2: Analyze traffic
– Q3: Study the question of thresholds and
triggering
– Q4: Identify requirements for hybrid
network and traffic engineering systems
20
Year 2 quarterly deliverables
• Year 2: Algorithm design and software
implementation
– Q1: Define interactions with DOEimplemented control-plane software
– Q2: Design and implement traffic
monitoring and characterization system
– Q3: Design algorithms and implement
hybrid traffic engineering system
– Q4: Design algorithms and implement
hybrid network engineering system
21
Year 3 quarterly deliverables
• Year 3: Prototyping on DOE-provided
testbed
– Q1: Test traffic monitoring and
characterization system
– Q2: Test hybrid traffic engineering
system
– Q3: Test hybrid network engineering
system
– Q4: Integration testing - Findings and
recommendations
22