Can Management Systems Leverage Self Organization? Goal Driven - - PDF document

1

Can Management Systems Leverage Self Organization?

Marc Pucci Chief Scientist Telcordia Technologies Applied Research marc@research.telcordia.com +1 732 699 2420 19 May 2005 Joint work with: James Alberi, Kong Cheng, Allen McIntosh SelfMan 2005 Panel 2

Goal Driven Adaptive Management of Converged Networks and Services

SelfMan – 2

Self-organization

– Engineer heuristics for goal-directed control of desired

emergent behavior

Distributed management

– Decentralize control loops, local knowledge only – Deterministic rules and policies

Centralized management

– One omniscient management system

Is there a middle ground? Where? What have we tried? What have we learned?

– In any complex system, grief is conserved – Mike O’Dell

2

SelfMan – 3

Challenges

Complex systems require multiple types of controllers

– Avoid destructive interference between controllers – How do we make sure multiple controllers work towards the greater good?

Local controllers optimize without global objectives

– E.g., Bandwidth load balancing in the face of DDoS attack – Tend to be local resource aware, not service aware – But can be self-organizing

Difficult to increase awareness in local controllers

– Usually impossible to modify proprietary systems – Too complex, will never keep up with growth of constraints – Wrong Approach? Still need a coarse-grain behavior modifier? – But is some amount of local interaction needed?

Too many rules

– Better knowledge representation – Adaptive weight setting, learning

SelfMan – 4

Where do we apply controls?

(e.g. Auto Attendant Meet-Me Conf Unified Messaging CallCenters) Wiretap Server Feature Svr (e.g IP Centrex) Application/Media Servers LNP Call Name 800 DB AIN SCP

Network DBs CCS Network Network OSSs

• Configuration
• Fault
• Performance

SNMP CORBA

Billing

AMADNS HTTP

PSTN Transport Network

Legacy Voice Mail T D M ( P R I , I S U P

• r

M F ) TCAP

PSTN Gateway Managed IP Network Session Border Controller Access

IAD Enterprise Consumer or SME Legacy PBX CAS/PRI PBX CAS or PRI PBX GW DNS/DHCP/ Radius Svrs SCP IP PBX Softswitch (Signaling GW MGC Feature Server Line Side Trunk Side) Softswitch

Intra or Inter Carrier Softswitches

Call Admission Control Settings Application Traffic Control Settings Internal MPLS Route Changes MPLS + Diffserv Ratio Settings Peer-to-Peer Network Routing IP Phone Codec Settings Access Control Bandwidth Settings Policing, Marking, Shaping Reduction of Permitted Burst Capacity

3

SelfMan – 5

Local Resource Controller example - Automatic Bandwidth Management

SelfMan – 6

MPLS Bandwidth Management

Closed Loop – Monitor – Analyze - Reconfigure

– Single Goal – Simple Objective function

• Minimize maximum load
• Minimize hop count over minimum possible hops

Use of Trending and Prediction Could be self organizing, adapting to load usage patterns But - Not Service Aware – all bandwidth users are equal

DATA Performance Measurement System VPN Loads Analysis Engine New LSPs Current Data Network Configuration, LSP Configuration, Performance Data Load Generators LSP Configuration Network Probes Control MPLS Network

4

SelfMan – 7

What happens when multiple applications have conflicting demands

• n the same supporting set of resources?

– Converged networks with voice, video, data – Differing demands on delay, jitter, bandwidth

No single, static solution will suffice

– Allow local controllers to operate in well-defined comfort zones – Be careful about allowing local controllers to conspire

E.g., VoIP manager dealing with MPLS bandwidth manager

– Invoke (passively or actively) higher level management when zones

are violated

– Don’t micro-manage

Additional considerations to manage:

– Changing priority of services and sessions within services – Profitability of services

Admit billable voice call rather than in-plan call Limit data capacity to allow high speed, high QoS video transmission Discard existing traffic to allow new applications to fit

SelfMan – 8

Experimental Telcordia Test Bed for Multi- Layer Converged Services Management

Optical Ring

ROADM ROADM ROADM ROADM Optical Manager

BoD Service Manager Bandwidth Manager (PECAN) IP/MPLS VoIP Service Manager Global Policy Manager (Arpeggio) UCSD BoD Service Client

• Independent Controllers manage private view of their world
• Free Running Controllers

–

Until exception occurs

• Rule driven interactions