CoLab session 1 deployment best practices and architecture
•Download as PPTX, PDF•
0 likes•863 views
The document discusses deployment best practices and architecture for next generation communications solutions. It recommends adopting a software-defined architecture that is interoperable, resilient to failures, and focuses on communications enablement. The session state server provides a modern and scalable presence solution with load sharing and redundancy. Deployment models include single or multi-server clusters to support different customer needs depending on size and locations. AudioCodes gateways help connect TDM networks to SIP applications, and stand-alone survivability provides call continuity in emergencies.
Report
Share
CoLab session 1 deployment best practices and architecture
- 1. Deployment Best Practices and Architecture Dave Deutschman, Alan Percy, George Niculae, Joegen Baclor Moderated by Martin Steinmann
- 2. Introduction to the Presenters Session 1: Deployment Best Practices and Architecture Joegen Baclor Dave Deutschman George Niculae Alan Percy
- 3. Best Practices for a Next Generation Solution Transition to all software and an IT application imply profound changes IT / all software solution managed as part of the IT stack Not vertically integrated: Standards based and interoperable Datacenter or cloud deployments, centrally managed Tolerant to hardware failures – resiliency in the SW architecture Focus on communications enablement of applications (single sign-on, user profiles, social graphs, presence, activity streams) 3
- 4. IT Best Practices Demand a New Architecture SIP Session Oriented Architecture - SSOA Internet Scale Communications • Native Session Initiation Protocol (SIP) architecture for a modern, highly scalable and resilient communications infrastructure • Based on a distributed and load-sharing network of SIP proxy servers • Geo-redundant redundancy reduces complexity and number of servers required • Peer-to-peer media for industry leading voice and video quality • Introducing presence based communications that facilitates social business integration and multi-media interactions > Scalable • Pushing communication and collaboration into the cloud > Resilient > Cost efficient
- 5. New Session State Server (SSS) Modern and scalable presence solution What is new in SSS ? Load-sharing redundancy and scale for line state presence and session state New technology from the ground up: • Fast and modern SIP Stack using reSIProcate • Multiple Watchers and State Agents • Load-balanced subscriptions via DNS/SRV • Shared Subscription State – no single point of failure • Centralized Event Queue 5
- 6. Session State Server (SSS) Architecture Fully redundant highly available deployment Failure cases addressed: Ability to detect failure of a node via real-time notification Standard SIP fail-over for transaction failures Load sharing using standard SIP DNS/SRV transaction spreading Bottle neck free fairly-queued task distribution 6
- 7. New in 4.6: Redundant Media Services Seamless global resiliency and scale openUC 4.4: Configuration and voicemail server tightly coupled – both reading data from local file system openUC 4.6: Improved voicemail REST API, configuration server retrieves data using voicemail REST API New voicemail architecture – abstract connector layer to allow pluggable storage types Decoupled voicemail + ability to read / write from external storages = foundation for redundant voicemail Efficient memory & disk usage, built in replication - GridFS storage connector Moving forward to cloud - Amazon S3 storage connector 7
- 10. Redundant Media Services 10
- 11. Redundant Media Services 11
- 12. Redundant Media Services 12
- 13. Redundant Media Services: How Does It Work? Seamless user experience with geo-redundancy Redundancy achieved at: Proxy level (DNS) Media Services level (DNS) Voicemail level via FreeSWITCH dial plan Distributed voicemail storage 13
- 14. Redundant Media Services 14
- 15. Redundant Media Services 15
- 16. Redundant Media Services 16
- 17. Redundant Media Services 17
- 18. Deployment Best Practices How to build a system 18
- 19. Health Provider with 10 Locations Design Considerations - Single data center - Reliable corporate network with QoS - Locations have different telco providers - Need for groups in the same location to use different caller IDs - 24 emergency hot lines - Limited remote worker support (IT) - Instant Messaging (IM) 19
- 20. Single Multi-Server Cluster Health Provider with 10 Locations 20
- 21. Deployment Model Benefits Single Multi-Server Cluster - Distributed services to support scalability and availability - Registration and call processing balanced across master and distributed proxy - ACD deployed on separate server - Centralized management of all locations - Local services - Network applications (DHCP and DNS) - PSTN access - Remote support through VPNs 21
- 22. Large Corporation with Many Locations Design Considerations - Multiple corporate data centers - Many locations with a large number of users - Reliable corporate network with QoS - Need for location survivability - Customer uses basic telephony services and voicemail - Large scale remote worker support 22
- 23. Multi-Server, Distributed Cluster Large Corporation with many Locations 23
- 24. Deployment Model Benefits Multi-Server, Distributed Cluster - Scalability and availability - Geo-redundancy for call processing service - Centralized management - Flexible PSTN options - Media Gateways - SIP Trunking - Scalable Remote Worker support through use of Session Border Controller 24
- 25. IT Best Practices: Multi-Vendor Hardware Giving you freedom and lower cost Disintegration of the vertical model: Certified and interoperable with a large set of devices Offering plug n’ play management and end-to-end support for a select set of vendor’s products Audiocodes gateways as the preferred choice for PSTN termination 25
- 26. AudioCodes Gateways and openUC Connecting TDM Networks to SIP Applications Key Benefits: • Plug & play configuration with openUC • Modularity, Scalability & Flexibility Mediant ™ 1000 • Interoperability Mediant ™ 800 mix-and-match Gateway/E-SBC for Business Cost effective • Survivability (SAS) MediaPack™ 124 Branch office Gateway/E-SBC With 24 FXS • Wide Availability MediaPack™ 11X with FXS/FXO interfaces via Distribution interfaces 26
- 27. SmartTAP Recording for openUC • Software based IP Centric all-in-one or distributed recording solution • Support from 8 - 300 concurrent sessions per server with expandability to grow with your organization by adding additional servers. • Scalable to 1000s of concurrent IP/TDM Channels • Designed for easy deployment, ease of use, easy maintenance 27
- 28. openUC Appliance with OSN3 Module Scalable/Modular Appliance for the SMB/SME Branch Office Create an appliance: openUC software on OSN3 PSTN module T1/E1 OSN3 or FXO Module AudioCodes Mediant 1000B SIP SIP Phones LAN Switch 28
- 29. eZuce openUC Appliance with Mediant 800 Up to 12 FXS ports for POTS OSN Module FXS WAN AudioCodes PoE LAN Mediant 800 SIP Phones 29
- 30. Introduction to Stand Alone Survivability At Typical Branch Office Digital SIP Phones AudioCodes E911 and Media Gateway or Survivability MSBG with SAS FXO Connection PSTN WAN 30
- 31. SAS - Registration Process Digital SIP Phones Phones register with MediaPack AudioCodes E911 and Media Gateway or Survivability MSBG with SAS FXO Connection PSTN Gateway forwards registrations to openUC WAN 31
- 32. SAS - Normal Dialing Mode Digital SIP Phones Phone user makes telephone call AudioCodes E911 and Media Gateway or Survivability MSBG with SAS FXO Connection PSTN Normal Mode: MediaPack forwards INVITES to openUC WAN 32
- 33. SAS - E911 Call Digital Phones user SIP Phones dials “911” MediaPack detects AudioCodes E911 and “911” and redirects Survivability Media Gateway or MSBG with SAS call to FXO port FXO Connection toward PSTN PSTN WAN 33
- 34. SAS - Emergency Mode Digital Phone user SIP Phones makes telephone call Gateway detects connection to network is down (keep Emergency Mode: MediaPack MP-112/FXO E911 and alive messages fail) and Media Gateway With SAS Gateway redirects Survivability activates all calls to FXO port FXO Connection Emergency Mode toward PSTN PSTN WAN 34
- 35. Simplicity matters It is measured in operating expense openUC is an order of magnitude simpler than previous generation solutions and therefore cheaper to operate and develop 35
- 36. Q&A • System architecture / SSOA • Deployment best practices • Geo-Redundancy • Gateways & phones • Network requirements and QoS • Setting up and managing a cluster 36