Just Jotted down my notes when reading on S4B Documentation…This could be a quick refresher to people well versed with Lync 2013
- FE Pool – min 1 server (3 recommended) max 12 – Max users 80,000 per pool
- Separate Backend (SQL) Pool for Enterprise (In Standard SQL Express is co-located and cannot be separated)
- Director Pool – max 10| Still not required but supported
- AlwaysOn Availability Groups & Failover Clustering (FCI) – now supported
- Director,Edge,Video Interop Server,Office Web Apps need separate pools
- Mediation, director, conferencing can be moved to separate pools if needed
Key Databases in Enterprise:
- Back End database
- Monitoring database
- Archiving database
- Persistent Chat database
- Persistent Chat compliance database
SQL Side Story
- Each SQL Instance can have only one instance of above…No sharing between pools please
- 2 Standard Edition Servers can be paired in 50:50 HA mode
- DR Pool is supported
- 20K Users per Edge Server
- HTTP Traffics needs Hardware load Balancer..DNS Load Balancing cannot be used – Applicable for WAC, Reverse Proxy
- Very Similar to LS2013
- OS: Minimum WS2012 | WS2008R2 Supported only for in-place upgrade situations
- Domain Function Level : WS 2003 minimum
- SQL: SQL Server 2008 R2 minimum- 64 bit only
- SQL Server mirroring and clustering not directly supported…Default is AlwaysOn in Topology Builder…Mentioned differently in another section(It says mirroring is default)!
- Silverlight still a pre-req …
- Only AlwaysOn Active/Passive is supported…NOT Active/active!
- Pre-Req Command for FE & : Add-WindowsFeature RSAT-ADDS, Web-Server, Web-Static-Content, Web-Default-Doc, Web-Http-Errors, Web-Asp-Net, Web-Net-Ext, Web-ISAPI-Ext, Web-ISAPI-Filter, Web-Http-Logging, Web-Log-Libraries, Web-Request-Monitor, Web-Http-Tracing, Web-Basic-Auth, Web-Windows-Auth, Web-Client-Auth, Web-Filtering, Web-Stat-Compression, Web-Dyn-Compression, NET-WCF-HTTP-Activation45, Web-Asp-Net45, Web-Mgmt-Tools, Web-Scripting-Tools, Web-Mgmt-Compat, Desktop-Experience, Telnet-Client
- Pre-Rq for directors: Add-WindowsFeature RSAT-ADDS, Web-Server, Web-Static-Content, Web-Default-Doc, Web-Http-Errors, Web-Asp-Net, Web-Net-Ext, Web-ISAPI-Ext, Web-ISAPI-Filter, Web-Http-Logging, Web-Log-Libraries, Web-Request-Monitor, Web-Http-Tracing, Web-Basic-Auth, Web-Windows-Auth, Web-Client-Auth, Web-Filtering, Web-Stat-Compression, NET-WCF-HTTP-Activation45, Web-Asp-Net45, Web-Scripting-Tools, Web-Mgmt-Compat, Desktop-Experience, Telnet-Client
- Please avoid ISA and LSP Softwares (Antivirus) …It will degrade multimedia performance
- Persistent Chat Pre-Req: MSMQ
- Default video codec is : H.264/MPEG-4 Part 10 Advanced Video Coding standard
- RTCP bandwidth:
- Audio 5kbps
- Video 10-15 kbps
- Audio :
- SILK – 54-100 kbps
- G.722 – 160-225 kbps
- H.264: 250 – 4000 kbps (min for static – 15kbps)
- PSTN – Only G.711 & RT Audio
- Qos using SDN Possible but requires additional Lync SDN Service Pool
- Lync Attendant 2010 still not updated
- Each Data has three replicas – 1 primary and 2 secondary
- Routing Group Level Quorum loss (SQL Side)
- Happens if less than 85% of the SQL servers are not running. Pool will not start in this condition!
- Pool size : 2 – 1 minimum,3 – 2 minimum,4 – 3 minimum,8-6 min,12 – 10 min
- Primary SQL used as Quorum…hence if primary SQL Fails…whole pool will fail!
- Pool Level Quorum Loss (FE Side)
- Assumes SQL is working fine
- Check at least 3 FE running when adding users
- When Pairing Pools for DR , ensure 3 FEs in DR also are working
- For Enterprise please run minimum 3 FEs. If less only needed run Standard Edition
- Pooling between Enterprise and Standard Edition not allowed….Careful… Topology Builder does not resist!
- Recovery time for pool failover and pool failback
- Not different from Lync 2013
- For pool failover and pool failback, the engineering target for recovery time objective (RTO) is 15-20 minutes. This is the time required for the failover to happen, after administrators have determined there was a disaster and started the failover procedures. It does not include the time for administrators to assess the situation and make a decision, nor does it include the time for users to sign in again after failover is complete.
- For pool failover and pool failback, the engineering target for recovery point objective (RPO) is 5 minutes. This represents the time measure of data that could be lost due to the disaster, due to replication latency of the Backup Service. For example, if a pool goes down at 10:00 A.M., and the RPO is 5 minutes, data written to the pool between 9:55 A.M. and 10:00 A.M .might not have replicated to the backup pool, and would be lost.
- All RTO and RPO numbers in this document assume that the two data centers are located within the same world region with high-speed, low-latency transport between the two sites. These numbers are measured for a pool with 40,000 concurrently active users and 200,000 users enabled for Lync with respect to a pre-defined user model where there is no backlog in data replication. They are subject to change based on performance testing and validation.
- The engineering targets for Central Management store failover are 5 minutes for recovery time objective (RTO) and 5 minutes for recovery point objective (RPO).
- Location Information Database (LIS) is not failed over!
- AOL, Yahoo, and Google Talk are no longer supported….or is it vice-versa!
- Third Party (Cisco TP Specifically) Video Solution Integration now available via the Video Interop(erability) Server aka VIS Server
- This has been tested for a very specific topology based on CUCM 10.5 and TC7.0 & Above but looks like other topologies exist but not directly supported by MS
- The following are the Video devices (VTCs)
- Cisco C40
- Cisco C60
- Cisco C90
- Cisco MX200
- Cisco MX300
- Cisco DX80
- Cisco EX60
- Cisco EX90
- Cisco SX20
- More will obviously get added later I find it intriguing that Jabber Video is missing in this list!
- From what I understand from the text is that Lync/S4B Clients would not directly communicate P2P with Cisco VTCs…Instead it will be moderated by VIS L
- Pool Failover Supported but for compatibility on Cisco side a workaround is needed where Same IP Address gets resolved to different Pool FQDNs
- Since the failover is not actively monitored from cisco side, active calls will be lost during failover
- Between the FE and VIS, failover is actively monitored every 10 seconds
- Roll Back to normal is not automatic and needs manual intervention
- VIS will work for both S4B and Lync 2013 clients (with latest CU updates)
- Packet Loss Recovery can be enabled by enabling FEC but will increase SIP Trunk Bandwidth by 50%!
- Regarding VIS Sizing, On the standard FE Specification, max 16 VTCs can be transcoded (Each actually have three streams each on simulcast with 720p, 360p and 180p)…Hence it is actually doing 48 streams!
- No integration with online services currently…
- Support for both exchange Online and Exchange 2013
- Persistent Chat can scale to 4 servers with each server handling 20K users simultaneously
- Max 150K configured users supported
- No topology change
- Stretched pool supported in both active/active (2+2 instances) mode and active/passive (4+4 instances) mode…but network has to be ‘high bandwidth, low latency’….what does that mean! I guess u can’t stretch across countries or even regions in large countries like in US
- No sizing calculator/guidance for virtual environments yet L
- Ethical walls still a Persistent Chat feature and not available on S4B FE L by default
PSTN Voice Services
- Enterprise voice is not the same as Call-Via-Work
- Remote Call Control not supported
- Enterprise Voice remains similar to Lync 2013
- Uses UCWA for Integration with PBX and hence features depend on the PBX’s UCWA Integration capabilities. Will be unique for each PBX
- E-911 and Malicious Call Trace, Mute/Unmute, Recording etc not supported by default
- Can Join Meeting Bridges but not perform ad-hoc conferencing
- VDI Not supported
- Direct calls by deskphone not logged by S4B
- CVW users must have DID on PBX
- CVW users consume EV License and needs EV enabled for each of them
- Delegation, Team Call and Response Group disabled for each CVW user
- Monitoring Capability Enhanced to provide ROI, Device Inventory, Helpdesk Grade Troubleshooting, System Troubleshooting Reports
- Please note that all these were possible earlier via third-party tools
- Lots of improvements here and lots of data points meaning the Database can get pretty huge too…So Plan for Separate SQL Instances
- Monitoring can be enabled/disabled at site level ..If that helps
- Again no calculator for sizing monitoring Storage / DBs
- Nothing much changed in archival front….My favourite still remains Exchange based archival