project co2: technical overview inder monga {imonga@nortelnetworks.c om} advanced technology cto...
Post on 24-Dec-2015
218 Views
Preview:
TRANSCRIPT
Project CO2: Technical Overview
• Inder Monga
• {imonga@nortelnetworks.com}
• Advanced Technology• CTO Office• Billerica MA
CO2 InterfacesCO2 middleware provides applications a service interface providing virtualized network view, dynamic provisioning, fault notification and performance monitoring
3rd party Service Creation and Mgmt
• Web Interface•HTTP
• Legacy IP/QoS•Classical RSVP
• (G)MPLS•CR-LDP•RSVP/TE
•UNI•ASTN UNI • MEF UNI
•Layer 2/RPR• SNMP• UNI• TL1
•CIM
PLUGINS
Service API
API to capture
And notify
events
CO2Intelligence and
Processing
VariousSignalingProtocols
Implemented
Policy Control and OAM
OAM API
Network provisioningEvents
• Manual• Traffic inspection• Application request
• Fault notification• Abstracted network
and performance view
Multiple Applications drive CO2
XML Messaging Service Bus
Storage ServerApplication
CO2 XML Application Interface
Core CO2 Service Intelligence
Signaling andManagement Layer
RadiologistWorkstationRequests data
Time reached to start Backup
Network Elements/Control Plane(OM3500, PP8600)/ASTN, GMPLS
Medical ServerApplication
GUIStation
Inside the CO2 box—a notional view
CO2-to-Control Plane scope
Networking stack, packet filters
CO2-to-CO2 scope
3rd party applications
CO2-to-OAM&P scope
PoliciesF
ilter
ed c
on
ten
t f
eed
Programming Model:Sample API invocations
Register (Authentication Info)
Register (Handle, SupportedServices)
Request (ServiceLevel, XferAmnt, Time, Priority)
Response (Status)
NetPoliciesDownload(PolicySchema)
SLAViolation (Latency)
IncTime (DeltaT)
Alert (Port X down)
Release (Handle)
App
lica
tion
CO
2 Softw
are
CO2 Features illustrated:
• Service Introspection
• Dynamic Policy control
• Abstracted service interface
• SLA Monitoring and Verification
• Error compartmentalization
CO2 Detailed Block Diagram
LEGEND
CO2 module
External Module
Internal API
External API
Cut-throughManager
(CTM)
AdvancedResv and
ToDScheduler
Network Provisioning API
Service Function call API Event interpretation API
Policy and OAM I/FService Creation and
Control
Service Creation API Policy and OAM API
Acceleration Services
ClientRegistation
AppSignaling/Messaging
UNI SessionManager
UNI Signalling
RPR CallManager
SNMP
GMPLS CallManager
GMPLSSignalling
Me
ch
an
ism
sA
pp
lic
ati
on
/P
ers
on
ali
tyS
ma
rts
QosManager
ApplicationPolicy
Database
COPS-PR
NT Extensions
SLAManager
Modulerepository/
maintenance
ModuleCreation
ModuleFunctioning
ContentForwarding
ContentRedirection
ACPServices
ACPResourceManager(ARM)
VPNController(VPNC)
CO2 ConfigManager
GUI
Application AAA
Personality -Service API
Service-Resource API
ApplicationPolicy
DecisionPoint
ServicePolicy
Database
ResourcePolicy
DecisionPoint
ServicePolicy
DecisionPoint
Se
rvic
eS
ma
rts
Re
so
urc
eS
ma
rts
MetaProvisioning Service
ResourcePolicy
Database
Resource InformationService
CO2 Inter-Domain ManagerCO2 RISDatabase
Resource Usage Allocation Resource Usage Optimization
Resource Usage Feedback
TL1
Service AAA
Resource AAA
POLICY
API
Monitoring&
Topology
Module Management
ModuleControl
A day in the life:Application driven provisioning
Application Software
(AppSoft)
Policy TOD
ConfigurationState of network
(bandwidth, VLANs)
OperationalState of network(errors, latency)
SmartB/w
Manager
2
3
4
5 7
8
6
1
1. AppSoft sends a request for bandwidth to CO22. After JMS and XML processing, the message is send
to smart bandwidth management module (SBM)3. SBM consults the policy engine with network
policies to figure out available bandwidth for this request
4. Policy engine looks at general and specific Time of Day (TOD) policies to calculate allowable bandwidth for this request
5. SBM consults the configuration state block to figure out total bandwidth already allocated for other ongoing requests, plus bandwidth available for this node
6. Check the operation state of the network to ensure proper performance level is met for the request
7. Send configuration commands with proper attributes to meet AppSoft request
8. Meta signaling API uses the right signaling blocks accomplish dynamic provisioning.
Indicates blocks that are ongoing activities regardless
of Application messagesM
eta
Sig
nali
ng
Lay
er
JMS
and
XM
L
AP
I
A day in the life:Application driven provisioning (contd.)
7. EMR requests Ethernet errors8. Network elements respond with Ethernet errors9. EMR updates state of network, constructs an JMS/XML message with the VLAN specific errors and delivers it to CO2’s JMS/XML module10. CO2’s JMS/XML module responds to FMon with network VLAN errors that the application is
interested in viewing.
1. FMon requests CO2 to monitor errors on VLAN
5002. After JMS/XML processing, the message is
delivered to CO2’s Error Monitoring and Reporting module (EMR)
3. EMR reviews network topology, looks up the policy configuration for that VLAN and downloads the appropriate policies
4. EMR looks up Service Discovery information5. EMR signals for SONET errors based on the
discovered service capabilities6. Network element responds with path and link
errors
A day in the life: Error monitoring
Fault MonitoringApplication
(FMon)
Policy TOD
ConfigurationState of network
(bandwidth, VLANs)
OperationalState of network(errors, latency)
ErrorMonitoring &
Reporting
23
4
6
9
7
1
Indicates blocks that are ongoing activities regardless
of Application messages
Met
a S
igna
ling
L
ayer
JMS
and
XM
L
AP
I10
5
3
7 8
8
9
1. StorageApp and VideoApp request a service affecting network resources.
2. After JMS/XML processing, the message is delivered to CO2’s Application Message Handler module (AMH)
3. AMH authenticates/authorizes and downloads the appropriate administrative policies relevant to the application request.
4. AMH uses the policy information to construct a detailed service request to the Smart BW Mgr.
5. QoM queries the Policy module for service policies relevant to the application requests.
A day in the life: Policy-based control
VideoApp
Policy
ConfigurationState of network
(bandwidth, VLANs)
OperationalState of network
(utilizationerrors, latency)
ApplicationMessageHandler
2
5
4
6
1
6. QoM queries the operational state of the network and configuration state of the network to get reachability and utilization information
7. QoM makes admission control decisions based on Applications service requirements, policies and its resource knowledgebase.
8. QoM provisions resources, if necessary, to meet StorageApp’s request
9. QoM sends a negative response to VideoApps request (lower priority than StorageApp).
10. Based on policies and network capabilities, CO2 can setup a new connection to grant VideoApps request if current capabilities cannot satisfy VideoApps request
Met
a S
igna
ling
L
ayer
JMS
and
XM
L
AP
I
9
3
7
StorageApp
QoS Manager
8
910
CO2 with CIM
• Management of CO2– CIM used opportunistically to
manage CO2 configuration• Utilize existing CIM policy
model
– Extend existing models for CO2 smarts management
• Management of CO2 Services
– Application interface to CO2 services currently via a proprietary XML API
– Investigating CIM for modeling CO2 service interface
• Requires extensive schema work to extend the CIM Core Model
3rd party Service Creation and Mgmt
API
Service API
ApplicationAPI
(CIM??)
CO2Intelligence and
Processing
VariousSignalingProtocols
Policy Control and OAM
OAM API (CIM)
• Manual• Traffic inspection• Application
request
• Event notification• Virtualized
network and performance view
• HTTP•IP/DiffServ• (G)MPLS•UNI•SNMP•TL1• CIM• …
CO2 Module Coding• Module & Code
– A module is an encapsulation of code providing a or more particular features/functions, e.g., SNMP, SLA, ToD, XML parsing
– CO2 module is implemented in Java + native codes
• Java package– Interface class: .java– Implementation class: .java– Makefile– Module profile: .mp– Module policy: .my– Doc: README– Binaries: .class– Unsigned Jar: .jar– Signed jar: .sjar
• Examples– Unsecured: examples/hello/server– Secured: examples/policy/source Network
Module
Module
Client
Network signaling
API invocation
App messaging
CO2
GUI
Native access
Device
CO2 CVS Source Directory Tree
testsstorage
grid client gui
skeletonprofiles
services
slam
qos
vpn
native
omninet
pp8600
om3500
common
protocols
resource
mib
networks metasignalling/drivers
runtime
jars
docs
contrib
tests
etc common
appmgr
policy
examples
templates
service
network
application
others
CO2
co2/
snmp
tl1
uni
scripts
config
com/nortelnetworks/co2/
Standards• Protocols:
– Messaging• JMS 1.1 (J2EE 1.4.1)• XML Parser, Xerces 1.4.4
– Signaling• RSVP RFC 3209• OIF UNI 1.0• Alteon NAAP• MEF UNI *• GMPLS *
– Policies• Policy Core Information Model -- Version 1 Specification (RFC 3060)• Policy Core Information Model (PCIM) Extensions (RFC 3460)
• Management: – SNMPv1, RFC 1157; SMIv1, RFC 1155– SNMPv2, RFC 1905; SMIv2, RFC 1902*– TL1– CIM*
• Grid:– OGSI– Globus/GT3
* In progress
Relevant Standards Bodies
• Global Grid Forum
• DMTF– CIM schemas for network devices and end-to-end services
• OIF– New UNIs
• IETF/IRTF– Policy, AAAs
• ITU– VPNs, (E)NNIs, GMPLS
• OASIS, W3C– Evolution of WS technologies
Related Work• GARA, DUROC
– Concept of Resource co-allocation, scheduler, advanced reservations leveraged in our work
– GPAN extends the reach of GARA/DUROC concepts
– Job Manager in GPAN refers to GRAM2 and its instances
• WS-Agreement– Services and resource lifetime-management and policy-based
negotiations between network domains
• GRAM/RSL/JSDL– Extend RSL2 to work with GPAN for network resources
– JSDL is new standard being discussed @GGF for job submissions
• Storage EMC, McData
• End-Systems IBM, HP-Labs
• Providers Allstream, Verizon
• Medical GE, Stryker, Cerner,
McKesson
• Conference Demos SuperComm, GGF9,
Telecom 2003, SC2003,
GlobusWorld 2004
• Testbed Winnipeg Health Authority
• External Funding DARPA
Engagements and Industry Validation
Demo 1: Application driven
Dynamic Bandwidth and QoS
In Billerica Lab!
Demo 2: A Globus-based Grid Infrastructure Negotiates
Ephemeral Optical Bandwidth Boost
Grid Proxy Architecture for Network Resources (GPAN)
• Enables Grid Resource Services to take advantage of existing network services
• The GPAN Grid middleware functionality includes:– Proxy for accepting Grid resource
requirements– Provider of information regarding network
resource availability/status– Co-existence and integration with GRAM2,
MDS– Support for RSL2 extensions featuring
network resource allocation capabilities– OGSI-services providing network resource
info & dynamic allocation capabilities– Abstract view and access to base network
services
Grid Applications
Network Elements
Grid Services
Network Services
GPAN
Grid Resources: general setup
B
DE
A Grid Domain:Virtual Organization
MDS Index
CA
• A Grid VO utilizes grid resources in Campus A-E• Service Providers (xSP) on MAN/WAN access networks peer together to provide required network services to the Grid VO.• Index services collects resource information from computing, storage resources in Campus A-E and xSP• Broker/metascheduler performs resource lookups and allocations of all grid resources for applications
Grid Overlay
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Storage
MFJSRIPS
MFJSRIPS
RMInfo
GRAM2RM
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Access
Network Service Overlay
Access
Broker/Metascheduler
Core Network xSPxSP
GPANProxyNIP
HostingEnvironment
application
Proxy architecture implements scalable resource services for networks
Computer
Network device
GPANGrid
Service
NetworkServices
• GPAN Grid Service– Provides a GRAM-2 instance in a network
– Extends RSL2 for network resources
– Supports resource discovery and info updates on the Grid
– Supports resource dynamic provisioning, optimization
– Resource Services such as GRAM talks to GPAN for network resource requests
– Grid clients and services use GPAN WSDL interface
GRAM2
Computing RM
GRAM2
VO Master CO2
GRAM2
Visual. RM
GRAM2
Storage RM
Broker/Metascheduler Application
Resource Management Flow
Network Service overlay
Network RM
MDS
Feeds not shown
Derived from © ANL Material
Network Resource Information using GPAN
B
DE
A Grid Domain:Virtual Organization
MDS Index
CA
Grid Overlay
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Storage
MFJSRIPS
MFJSRIPS
RMInfo
GRAM2RM
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Access
Network Service Overlay
Access
Core Network xSPxSP
GPANProxyNIP
• GPAN provides network info to MDS/Index– Proxy for network resource allocation status and updates
• Network Info Provider (NIP) aggregates resource discovery and status updates− Based on virtual network topology related to the VO
HostingEnvironment
application
Broker/Metascheduler
Network Resource Allocation using GPAN
B
DE
A Grid Domain:Virtual Organization
MDS Index
CA
Grid Overlay
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Storage
MFJSRIPS
MFJSRIPS
RMInfo
GRAM2RM
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Access
Network Service Overlay
Access
Core Network xSPxSP
GPANProxyNIP
1) Application requests broker/metascheduler for job services and resources
2) Broker/metascheduler generates RSL2 for resource allocation requests after consulting MDS/Index
3) xSPs co-ordinate to allocate requested resources
1
2
2
RSL2
RSL2
3
HostingEnvironment
application
Broker/Metascheduler
CO2 Grid with OGSI
Grid Applications
Network Info ProviderCO2 Grid Proxy
XML messaging
App Messaging
CO2 Grid HandlerRISService
Creation
CO2 GridPersonality
CO2 Core Platform CO2 Domain Mgr
Network Resources
CO2
OGSI
OGSA GridPlatform
OGSA Services
Network
CO2 Core Middleware
App. AResources Ethernet
Switch
EthernetSwitch
Photonicswitch
EthernetSwitch
Optical Control Plane
NetworkProvisioning
Services
GPAN Proxy
App. BResources
App. AResources
App. BResources
“A Globus-based Grid Infrastructure Negotiates Ephemeral Optical Bandwidth Boost”
Internet
Optical bypass
RSL2
• Two applications use Grid FTP for communicating large sets of data
• Grid FTP provides data movement requirements and constraints to GPAN
• GPAN Proxy module translates Grid requirements to appropriate network resource allocation
• GPAN Proxy module works with Network provisioning services to allocate optical by-pass as shown.
Demo 2: Default Provisioned Bandwidth for Clients
Demo 2: Dynamic provisioning of EphemeralOptical Bypass Circuit
Demo 2: GPAN Information Pane
Demo 3: DWDM-RAM DARPA Sponsored Research for Data Intensive Service-on-Demand Advanced Optical Networks
The Data Intensive App Challenge: Emerging data intensive applications in the field of HEP,astro-physics, astronomy, bioinformatics, computational chemistry, etc., require extremely high performance andlong term data flows, scalability for huge data volume,global reach, adjustability to unpredictable traffic behavior, and integration with multiple Grid resources.
Response: DWDM-RAM An architecture for data intensive Grids enabled by next generation dynamic optical networks, incorporating new methods for lightpath provisioning. DWDM-RAM is designed to meet the networking challenges of extremely large scale Grid applications. Traditional network infrastructure cannot meet these demands, especially, requirements for intensive data flows
Data-Intensive Applications
DWDM-RAM
Abundant Optical Bandwidth
PBs Storage
Tbs on single fiber strand
Optical Abundant Bandwidth Meets Grid
Optical Control Network
Optical Control Network
Network Service Request
Data Transmission Plane
OmniNet Control PlaneODIN
UNI-N
ODIN
UNI-N
Connection Control
L3 router
L2 switch
Data storageswitch
DataPath
Control
DataPath Control
DATA GRID SERVICE PLANEDATA GRID SERVICE PLANE
1 n
1
n
1
n
DataPath
DataCenter
ServiceControl
ServiceControl
NETWORK SERVICE PLANENETWORK SERVICE PLANE
GRID Service Request
DataCenter
DWDM-RAM Service Control Architecture
Data Management ServicesOGSA/OGSI compliant, capable of receiving and understanding application requests, have complete knowledge of network resources, transmit signals to intelligent middleware, understand communications from Grid infrastructure, adjust to changing requirements, understands edge resources, on-demand or scheduled processing, support various models for scheduling, priority setting, event synchronization
Intelligent Middleware for Adaptive Optical NetworkingOGSA/OGSI compliant, integrated with Globus, receives requests from data services and applications, knowledgeable about Grid resources, has complete understanding of dynamic lightpath provisioning, communicates to optical network services layer, can be integrated with GRAM for co-management, architecture is flexible and extensible
Dynamic Lightpath Provisioning ServicesOptical Dynamic Intelligent Networking (ODIN), OGSA/OGSI compliant, receives requests from middleware services, knowledgeable about optical network resources, provides dynamic lightpath provisioning, communicates to optical network protocol layer, precise wavelength control, intradomain as well as interdomain, contains mechanisms for extending lightpaths through E-Paths - electronic paths, incorporates specialized signaling, utilizes IETF – GMPLS for provisioning, new photonic protocols
DWDM-RAM Components
DataCenter
1
n
1
n
DataCenter
Data-Intensive Applications
Dynamic Lambda, Optical Burst, etc., Grid services
Dynamic Optical Network OMNInet
DataTransfer Service
Basic NetworkResource
Service
NetworkResource Scheduler
Network Resource Service
DataHandlerService
Information S
erviceApplication MiddlewareLayer
Network ResourceMiddlewareLayer
Connectivity and Fabric Layers
OGSI-ification API
NRS Grid Service API
DTS API
Optical path control
DWDM-RAM Architecture
4x10GE
Northwestern U
OpticalSwitchingPlatform
Passport8600
ApplicationCluster
• A four-node multi-site optical metro testbed network in Chicago -- the first 10GE service trial!• A test bed for all-optical switching and advanced high-speed services• OMNInet testbed Partners: SBC, Nortel, iCAIR at Northwestern, EVL, CANARIE, ANL
ApplicationCluster
OpticalSwitchingPlatform
Passport8600
4x10GE
StarLight
OPTera Metro5200
ApplicationCluster
OpticalSwitchingPlatform
Passport8600
4x10GE8x1GE
UIC
CA*net3--Chicago
OpticalSwitchingPlatform
Passport8600
Closed loop
4x10GE8x1GE
8x1GE
8x1GELoop
OMNInet Core Nodes
The DWDM-RAM architecture identifies two distinct planes over the dynamic underlying optical network: 1) the Data Grid Plane that speaks for the diverse requirements of a data-
intensive application by providing generic data-intensive interfaces and services and
2) the Network Grid Plane that marshals the raw bandwidth of the underlying optical network into network services, within the OGSI framework, and that matches the complex requirements specified by the Data Grid Plane.
At the application middleware layer, the Data Transfer Service (DTS) presents an interface between the system and an application. It receives high-level client requests, policy-and-access filtered, to transfer specific named blocks of data with specific advance scheduling constraints.
The network resource middleware layer consists of three services: the Data Handler Service (DHS), the Network Resource Service (NRS) and the Dynamic Lambda Grid Service (DLGS). Services of this layer initiate and control sharing of resources.
DWDM-RAM Architecture
Application
Fabric“Controlling things locally”: Access to, & control of, resources
Connectivity
“Talking to things”: communication (Internet protocols) & security
Resource
“Sharing single resources”: negotiating access, controlling use
Collective
“Coordinating multiple resources”: ubiquitous infrastructure services, app-specific distributed services
Data TransferService
NetworkResource
Service
Data Path ControlService
Layered DWDM-RAM Layered Grid
’s
Application
Optical ControlPlane
Application MiddlewareLayer
Network ResourceMiddlewareLayer
Connectivity &Fabric Layer
OGSI-ification API
NRS Grid Service API
DTS API
DWDM-RAM vs. Layered Grid Architecture
Network and Data Transfers scheduled• Data Management schedule coordinates network, retrieval, and sourcing services (using their schedulers)• Scheduled data resource reservation service (“Provide 2 TB storage between 14:00 and 18:00 tomorrow”)
Network Management has own schedule• Variety of request models:
• Fixed – at a specific time, for specific duration• Under-constrained – e.g. ASAP, or within a window
Auto-rescheduling for optimization• Facilitated by under-constrained requests• Data Management reschedules for its own requests or on request of Network Management
Design for Scheduling
• Request for 1/2 hour between 4:00 and 5:30 on Segment D granted to User W at 4:00
• New request from User X for same segment for 1 hour between 3:30 and 5:00
• Reschedule user W to 4:30; user X to 3:30. Everyone is happy.
Route allocated for a time slot; new request comes in; 1st route can be rescheduled for a later slot within window to accommodate new request
4:30 5:00 5:304:003:30
W
4:30 5:00 5:304:003:30
X
4:30 5:00 5:304:003:30
WX
Example: Lightpath Scheduling
0.5s 3.6s 0.5s 174s 0.3s 11s
OD
IN S
erve
r P
roce
ssin
g
File
tra
nsfe
r do
ne,
path
re
leas
ed
File
tra
nsfe
r re
ques
t ar
rives
Pat
h D
eallo
cati
on
req
ues
t
Dat
a T
ran
sfer
20 G
B
Pat
h ID
re
turn
ed
OD
IN S
erve
r P
roce
ssin
g
Pat
h A
lloca
tio
n
req
ues
t
25s
Net
wo
rk
reco
nfi
gu
rati
on
0.14sT
ran
spo
rt
setu
p t
ime
End-to-end Transfer Time
20GB File TransferSet up: 29.7s
Transfer: 174sTear down: 11.3s
20GB File Transfer
Conclusions
• Adaptivity requirements prompt us to re-define an application’s experience with the network
• We see the roll-out of on-demand solutions end-to-end as a multi-phase and multi-party concerted effort inclusive of network providers
• The CO2 framework is adaptivity middleware wherein we champion new contracts between applications and the network
• Storage EMC, McData
• End-Systems IBM, HP-Labs
• Providers Allstream, Verizon
• Medical GE, Stryker, Cerner,
McKesson
• Conference Demos SuperComm, GGF9,
Telecom 2003, SC2003,
GlobusWorld 2004
• Testbed Winnipeg Health Authority
• External Funding DARPA
Engagements and Industry Validation
A Strong Record in Light Paths Dynamic Provisioning
• OMNInet, the 1st wavelength-switched all-optical metro net, with end-to-end λ svcs established within seconds (2001)
– no SONET, no routers, no point-and-click
• SuperComm 2001, the 1st display of applications allocating lightpaths via OIF UNI
• GGF9 (Oct 2003), the 1st display of OGSI-fied λ services over a wavelength-switched all-optical network (“DWDM-RAM”)
• SC03, + time scheduling of under-specified client requests
• GlobusWORLD04, + integration with GT3, use by GT3 clients
• 3 journal submissions submitted for publication over 2H03 (JSAC, CCGrid, JOGC)
Some key folks checking us out at our CO2+Grid booth, GlobusWORLD ‘04, Jan ‘04
Ian Foster and Carl Kesselman, co-inventors of the Grid (2nd,5th from the left) Larry Smarr of OptIPuter fame (6th and last from the left)
Franco, Tal, and Inder (1th, 3rd, and 4th from the left)
CO2 Value Propositions• Topology discovery
– Automatic network discovery and storage to network linkage enables reduced management and outages costs
– Logical Topology discovery allows tracking network related application resource utilization
• Compartmentalization of faults – Expedited fault isolation and resolution– Reduced TCO leading to customer satisfaction and further investment in similar products
• Unified console for storage and network management– Collate application errors with network errors– Reduce IT management costs by managing performance of applications and corresponding network topology via
one interface
• Performance monitoring– Performance measurement of latency, packet jitter, bit error rate and other network related sensitivities as relating
to the application– Increased application uptime, response to potential problems before they become noticeable– IT organization can devote less time for active network monitoring and problem resolution
• Cross-departmental charging– Pay per network service usage– IT organization can identify user needs, justify expenses and plan for growth
• Absorbs churn from network upgrades– Seamless upgrade of network capabilities without affecting unified management
CO2 Value Propositions contd.
• Policy based Admission control– Improve application response time by restricting access to lower priority traffic during peak demands
– Translates application QoS requirements to network service provisioning
• Bandwidth efficiencies– Bandwidth savings through consolidating and managing multiple streams of storage/LAN traffic over a
common infrastructure
– Provides the IT department more efficient and longer use of current resources
• Configuration automation– Provides a web services interface for data center on-demand applications to access network services
through CO2
• Access to layered network services– Provides choice of service access at Layer 1 (SONET), Layer 2 (Ethernet/ATM/FR) and/or Layer 3 (IP)
depending on SLA parameters requested
• Performance smoothing via congestion avoidance– Eliminate erratic application performance during congestion by utilizing proactive bandwidth management
– Ensure applications receiving a consistent quality of service
• Application-Network QoS guarantees– Translates application QoS requirements to network service provisioning
Is CO2 a good fit for my network?Is CO2 a good fit for my network?
Does the network administrator allow access to network provisioning plane?
no yes
Do I want to use network’s standard provisioning plane?
yesno
CO2 Value Add Smart BW MgmtDynamic VPNsDynamic QoS
Using CO2’sframeworks,
write the plug-in code for a
particular signalingprotocol. Selected
unmodified built-in CO2 services work still,you can write new ones
no
yes
RSVP, GMPLS,
SNMP, TL1, UNI,
no
Does the network expose management information?
Does the network contain L2-L7 packet inspection boxes?
yes
CO2 Value Add Error Isolation
Performance MonitoringTopology SummarySLA Monitoring &
Verification
CO2 Value Add Accelerated processing,
content Insertion
Agile Network
C O2
Most Value-add
Limited Value-
add
Limited Value- add-
Brand X Nortel
Brand Y
Nortel
END
START
Virtual Organization
Amsterdam
Dwingelo Utrecht
Eindhoven
Amsterdam
Grid Overlay
CO2 Overlay
CO2+DiffServ
CO2+MPLSCore Network
Access
CO2
AccessCO2
Access
CO2
Access
CO2
Access
CO2
Example: CO2 brokering λs for a Globus application - general setup
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Storage
MFJSRIPS
MFJSRIPS
TBDInfo
GRAM2TBD
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Acronyms:•MMJFS: Master Managed Job Factory Service•MJFS: Managed Job Factory Service•MJS: Managed Job Service
•RIPS: Resource Info Provider Service•MDS: Monitoring and Discovery Service•GRAM: Grid Resource Allocation Management
HostingEnvironment
application
MDS Index
VO’s Meta Scheduler
Virtual Organization
Amsterdam
Dwingelo Utrecht
Eindhoven
Amsterdam
Grid Overlay
CO2 Overlay
CO2+DiffServ
CO2+MPLSCore Network
Access
CO2
Access
Access
CO2
Access
CO2
Access
CO2
Example: CO2 brokering λs for a Globus application - Resource discovery
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Storage
MFJSRIPS
MFJSRIPS
TBDInfo
GRAM2TBD
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Acronyms:•MMJFS: Master Managed Job Factory Service•MJFS: Managed Job Factory Service•MJS: Managed Job Service
•RIPS: Resource Info Provider Service•MDS: Monitoring and Discovery Service•GRAM: Grid Resource Allocation Management
HostingEnvironment
application
MDS Index
VO’s Meta Scheduler
VO’s Master CO2
Virtual Organization
Amsterdam
Dwingelo Utrecht
Eindhoven
Amsterdam
Grid Overlay
CO2 Overlay
CO2+DiffServ
CO2+MPLSCore Network
Access
CO2
Access
Access
CO2
Access
CO2
Access
CO2
Example: CO2 brokering λs for a Globus application - Resource allocation
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Storage
MFJSRIPS
MFJSRIPS
TBDInfo
GRAM2TBD
Computing
GRAM2MMJFS
MFJSRIPS
MFJSRIPS
MJSRIPS
Acronyms:•MMJFS: Master Managed Job Factory Service•MJFS: Managed Job Factory Service•MJS: Managed Job Service
•RIPS: Resource Info Provider Service•MDS: Monitoring and Discovery Service•GRAM: Grid Resource Allocation Management
HostingEnvironment
application
MDS Index
VO’s Meta Scheduler
VO’s Master CO2
Scheduling Example - Reroute • Request for 1 hour between nodes A and B between 7:00 and
8:30 is granted using Segment X (and other segments) is granted for 7:00
• New request for 2 hours between nodes C and D between 7:00 and 9:30 This route needs to use Segment E to be satisfied
• Reroute the first request to take another path thru the topology to free up Segment E for the 2nd request. Everyone is happy
A
D
B
C
X7:00-8:00
A
D
B
C
X7:00-8:00
Y
Route allocated; new request comes in for a segment in use; 1st route can be altered to use different path to allow 2nd to also be serviced in its time window
top related