OOF Architecture Review R 3 Date Aug 14

  • Slides: 10
Download presentation
OOF Architecture Review (R 3) Date Aug 14, 2018 Shankaranarayanan Puzhavakath Narayanan

OOF Architecture Review (R 3) Date Aug 14, 2018 Shankaranarayanan Puzhavakath Narayanan

ONAP Optimization Framework (R 3) ONAP Optimization Framework (OOF) provides a declarative, policy-driven approach

ONAP Optimization Framework (R 3) ONAP Optimization Framework (OOF) provides a declarative, policy-driven approach for creating and running a wide variety of optimization applications like Homing/Placement, PCI optimization, etc. Policy Models OOF SDNC CM Portal simulator SO PCI Handler MS Route CMSO HAS PCI Optimizer SDNC API CMSO Optimizer Route Optimizer A&AI HAS CMSO OSDF Homing/Placement Optimizer (HAS) Multi Cloud Policy PCI Actual policies Dependencies on other ONAP Projects MUSIC 2

OOF Repos in R 3 • OSDF: Optimization Service Design Framework • HAS: Homing

OOF Repos in R 3 • OSDF: Optimization Service Design Framework • HAS: Homing and Allocation Service • CMSO: Change Management Schedule Optimizer • Seed code being up-streamed in R 3 3

IM/DM Alignments - Service and Resource Info, from: AAI HPA Flavors/Capabilities/Capacity Info, from :

IM/DM Alignments - Service and Resource Info, from: AAI HPA Flavors/Capabilities/Capacity Info, from : AAI Policy Models (Homing, PCI) from: Policy Infrastructure Metrics Info (capacity), from: Multi. Cloud agnostic Intent Info, from: Multi. Cloud PCI configuration data (not yet a part of SDC model) 4

Exported API Service Orchestrator: - https: //wiki. onap. org/pages/viewpage. action? page. Id=25435066 Change Management

Exported API Service Orchestrator: - https: //wiki. onap. org/pages/viewpage. action? page. Id=25435066 Change Management Simulator: - https: //wiki. onap. org/display/DW/CM+API+Specifications PCI Handler MS - https: //wiki. onap. org/display/DW/PCI+Optimization+API SDNC - https: //wiki. onap. org/display/DW/Route+Optimization 5

S 3 P Updates Optimization Service Design Framework (OSDF) Performance Stability Homing and Allocation

S 3 P Updates Optimization Service Design Framework (OSDF) Performance Stability Homing and Allocation Service (HAS) Level 1, develop plans for performance improvement Level 1 Resiliency Level 2; OSDF is stateless; clients retry requests if OSDF container crashes or times-out. Clients manage timeouts. Security CII Passing Badging – No critical security and licensing issues Increase Code coverage from >50% to >60% AAF based authentication; encrypted communication between components Scalability Level 1 Level 3; HAS provides Active/Active resiliency across multiple sites Level 2 Manageability Level 1; with progress on adherence to Logging v 1. 2 spec Usability Level 1 6

Backup 7

Backup 7

Optimization Framework – R 3 (1/4) • • Service Orchestrator (SO) Change Management Portal

Optimization Framework – R 3 (1/4) • • Service Orchestrator (SO) Change Management Portal SDNC PCI Handler MS Optimization Framework (OOF) • • • Policy Interface Inventory Service Interface Service/Policy Models Interface Infrastructure Metrics Interface MUSIC REST API interface Config data interface, SDNC(R) Definition: - ONAP Optimization Framework (OOF) enables easy creation of optimization applications by leveraging a policy-driven, declarative approach, that supports different specializations (applications) with specific domain and optimization needs. - Current scopes include (but are not limited to) VNF placement support via Homing and Allocation Service (HAS) with different use cases, change management scheduling service, Route Optimization, and PCI Optimization. Provided Interfaces: - Placement Optimization Interface - Provides functionality for the Service Orchestrator to enable optimized homing/placement of Services - PCI Optimization Interface - Provides functionality for PCI Allocation to minimize conflicts and confusion - Route Optimization Interface - Provides functionality for path/route optimization given a topology - Change Management Portal Interface - Provides functionality for calculating schedules that satisfy time constraints and conflicts [CM scheduling use case] Consumed Interfaces: - Policy Interface, from: Policy - Inventory Service Interface, from: Available and Active Inventory - Service/Policy Models Interface, from: SDC - Infrastructure Metrics Interface, from: Multi. Cloud - MUSIC REST API Interface, from : MUSIC - Config data interface for PCI optimization, from : SDNC(R) Consumed Models: - Policy Models (homing, PCI), from: SDC - Infrastructure Metrics Models, from: Multi. Cloud 8

OSDF Architecture 9

OSDF Architecture 9

HAS Architecture SO HAS architecture supports Active/Active mode Homing solution to SO (after reservations,

HAS Architecture SO HAS architecture supports Active/Active mode Homing solution to SO (after reservations, if required) Homing Request from SO based on SDC models + runtime parameters HOMING-API OSDF (policy translation, validation) A&AI MC Fetch homing policies for request Policy Homing Request with policy constraints HAS-API Data/Information plugins HAS-Data HAS Component (process) OSDF Component In memory cache candidates Fetch initial ers Run static filt Ite for rative sol l vin y fetc gc on h dat Cal str l Re ain a ser ts vat ion API s HAS-Controller Optimization Algorithm plugins mplate, r eady to b HAS-Optimizer/Solver Asynchronous/polling Intra HAS component messaging Homing t e HAS-Reservation Constraint Model plugins Homing solution (after reservations) e solved *MUSIC DB Homing solution (without reservations) * HAS uses a simple abstraction of gets/puts, and is DB agnostic 10