Naseem Asif Holger
0 likes364 views
The document discusses designing highly available voice over internet protocol (VOIP) systems. It covers establishing system requirements, determining deployment configurations, and mapping middleware capabilities to solutions. Key aspects include essential and management services, hardware platforms, functional blocks, and redundancy policies like N+1 models. The GoAhead software platform enables developing and deploying available network elements using open standards.
![Thank you! Dr. Asif Naseem, GoAhead Software President and COO [email_address] 425.453.1900](https://image.slidesharecdn.com/naseem-asif-holger-1206220518501953-2/85/Naseem-Asif-Holger-20-320.jpg)
Naseem Asif Holger
- 1. Designing Highly Available VOIP Systems Dr. Asif Naseem President and COO, GoAhead Software, Inc.
- 2. Designing a VOIP Network Element Establish system requirements Determine deployment configuration Mapping middleware capabilities to the desired solution
- 3. Establishing System Requirements Essential Services Systems Management Availability Management Platform Management Foundational Kernel Application Services
- 4. Systems Management Services Enable Creation of Internal management functionality External management functionality EMS interface SNMP Agent HPI MIB Alarm Management Log AMF MIB Information Model Mgmt Notification Systems Management Services
- 5. Availability Management Services Center piece of HA systems HA management Object CRUD Administration operations Dependency mapping Redundancy policies Availability Management Services Availability Management Framework
- 6. Platform Management Services Enables management of platform capabilities Resource discovery Monitor and control Population of system model Static/dynamic configuration Hardware agnostic Platform Management Services Platform Resource Management Service Hotswap Management Service
- 7. Foundational Services Foundational Services provide A variety of functionality that system developers can use to build highly available systems Foundational Services Cluster Management Distributed Messaging EJ Console
- 8. Kernel Services Kernel provides Small, reliable foundation Cross platform Abstraction of platform specific capabilities into generic platform independent capabilities Kernel Services Hardware abstraction Loading Memory Mgmt
- 9. Application Services Applications services Services intended for application developers Application Services Resource Virtual IP Messaging Management Datastore Checkpointing Event Cluster Membership
- 10. Putting It All Together Hardware Platform Running Standard Operating System Kernel Foundation Services SNMP Agent HPI MIB Alarm Management Log AMF MIB Information Model Mgmt Notification Platform Resource Management Service Resource Virtual IP Messaging Management Datastore Checkpointing Event Cluster Membership Systems Management Services Platform Management Services Availability Management Services Application Services Hotswap Management Service Availability Management Framework Cluster Management Distributed Messaging Action EJ Console Webs Base Applications
- 11. Determine Deployment Configuration Picking an appropriate hardware platform – HPI enabled Hardware Shelf manager Switch fabric Blades for various purposes Establishing appropriate functional blocks Applying desired redundancy policies
- 12. Functional Blocks Operations, administration and maintenance (OA&M) System manger Monitors system state Provides management interface required by external management entities Call control Voice processing services Communicates with OA&M elements Communicates with Media Control element Media control Manages switching configuration Communicates with active OA&M element
- 13. Redundancy Policies Component: a set of software and/or hardware resources SU: Aggregates a set of components to provide a particular service SG: contains one or more SUs that participates in a redundancy model to provide service availability Service Group Service Unit 1 . . Component 1 Component 1 Service Unit 2 . . Component 1 Component 1 Active Standby
- 14. 2N Redundancy Model 1 + 1 1 + 1 At most one SU has the active HA state At most one SU has the standby HA state Service Group 1 Service Unit 1 . . C1 Cn Service Unit 2 . . C1 Cn Active Standby Service Group 2 Service Unit 1 . . C1 Cn Service Unit 2 . . C1 Cn Active Standby
- 16. Mapping Middleware Capabilities: OA&M Node Manager Capable nodes OA&M Node SRP IMM Manager PRMS Notification Log SNMP Agent AMF Manager AMF Client Cluster Management Distributed Messaging Information Model System Model
- 17. Mapping Middleware Capabilities: Call Control Node Client-only nodes Call Control Node SRP Event Checkpointing Cluster Management Distributed Messaging AMF Client
- 18. Mapping middleware capabilities: Media Control Node Client-only nodes Media Control Node SRP Messaging Cluster Membership Cluster Management Distributed Messaging AMF Client
- 19. Key Messages It is practical and feasible to put together commercially deployable highly available VOIP systems using COTS components SA Forum standards, among others, are helping this cause GoAhead has field proven software platform that is being used to develop and deploy a variety of highly available network elements
- 20. Thank you! Dr. Asif Naseem, GoAhead Software President and COO [email_address] 425.453.1900