TenTips_GlenBrumpton_NUCUG_Feb_1.1_Cleaned

Northern UC User Group10 “Things” to watch for to help deliver a successful Lync deployment

Presenter : Glen Brumpton

Web : northernucusergroup.co.uk Twitter @northernucuserg

About me:

Unified communications platform

• Lead UC Architect for Capita IT Services

• Role solely focuses on Microsoft Lync and Enterprise Voice

• Pre-Sales, Architecture Design and Solution Delivery

• 21 Years consultancy experience working for Global SI’s and Telco’s

• Previously : Technology Consultant – KCOM & Affiniti

• Formerly : Technical Consultant - Capgemini

• Lync Voice P-TSP on behalf of Microsoft

• Partner Technology Solutions Professional

• Extension of the Microsoft Lync TSP team

• From the North East of England

• Marske By The Sea

• 3 Children

• Keen Motorcycle Rider

• Sometimes follow Middlesbrough Football ClubWeb : northernucusergroup.co.uk - Twitter @northernucuserg

#1 Group Call PickupChallenge the use of Group Call Pickup

• Supported in Lync 2013 since CU1, but unwieldy for both users and administrators

• Use SEFAUtil to enable a user for Group Call Pickup and assign a pickup number

• Call Park application – New-CSCallParkOrbit cmdlet to create call pickup number ranges in the call park orbit table and assign the pickup ranges the type GroupPickup

• Response Group / Delegate Calls / Team Calls / SimRing Calls cannot be picked up using Group Call Pickup

Before Lync

• Sue calls Bob, Bob isn’t at his desk (Sue wouldn’t have know this as there’s no presence )

• Paddy can hear Bob’s phone ringing and uses a feature code **3 / *31 to answer Bob’s phone

• Paddy leaves a message on a “post-it note” for Bob to call Sue when he’s back.

After Lync (Without Group Call Pickup deployed)

• Sue calls Bob, Bob isn’t at his desk (Sue knows this as he’s showing away, but decides to call him anyway)

• Scenario 1 - Bob has enabled SimRing to his mobile, Sue’s call is forwarded to Bob’s mobile and the call is connected

• Scenario 2 – Bob hasn’t enabled SimRing and instead Sue’s call is forwarded to his voicemail

Web : northernucusergroup.co.uk - Twitter @northernucuserg

#1 Group Call PickupPrior to a deployment

• Identify “function champions” – i.e. a user from Finance, a user from Payroll, a user from Admin etc.

• Work with the function champions to help them distinguish between calling a person and calling a function

• If I want to speak to the “Finance Department” I’ll call the relevant RSG

• An available agent of the Finance Department RSG will answer the call

• If I want to speak to Bob I’ll call Bob

• If Bob’s not available the inbound call will be handled however Bob has chosen the call to be handled

• There’s no need / business benefit to leaving messages on post-it notes for an individuals missed call

Once a legacy PBX users mindset has been changed to one of a UC “way of working” there shouldn’t be a business need for Group Call Pickup


#2 QuorumPool Quorum

• For a pool that is running, at least 50% of the FE servers need to be online to achieve quorum state

• For a pool starting from a cold boot 85% of the FE servers need to be online to achieve quorum state and for FE services to start


# of FE Servers in the pool # of FE Servers that must be running for pool to be functional

1-2 1

3-4 2

5-6 3

7-8 4

9-10 5

11-12 6

#2 QuorumReplica Set Quorum

• Not related to Pool Quorum, Used by Routing Groups

• Each Routing Group is assigned a Primary, Secondary and Tertiary Front End Server

• Windows Fabric allocates the copies of the routing groups to different servers, ensuring a single server only contains one copy of a routing group

• Windows Fabric also relocates routing groups if a Front End Server goes offline containing a primary copy

• Users are automatically assigned to a Routing Group when the user account is provisioned

• Lync 2010 hash algorithm no longer used

• Routing Group information is held in AD (msRTCSIP-UserRoutingGroupId)

• Routing Groups must maintain Quorum, if more than 1 server from a Routing Group is lost at the same time service will be lost to users belonging to the failed Routing Group

• SBA’s / SBS’s get their own Routing Group however, all users assigned to the SBA/SBS will be serviced by this Routing Group


#3 VirtualisationHypervisor

• Microsoft SVVP Program lists supported hypervisors for Lync

• Disable Hyper-Threading on the Hosts

• Dedicate CPU Cores to Virtual Machines

• Reserve RAM to Virtual Machines

• Use 10Gb interfaces if possible

• VMware DRS / vMotion – Hyper-V LiveMigration not supported

• VMware SRM not supported, use Pool Pairing for DR in Lync

Virtual Server Placement

• 1 Front-End Virtual server per host

• Split the SQL Virtual Servers across multiple hosts

• In a Pool Quorum voting scenario only the principal SQL server is used to vote


#4 Network CurveballsWAN Accelerators / Optimizers

• Riverbed Steelheads / Cisco WAAS

• Mitigate the impact of low network bandwidth and delay

• Data compression

• Caching

• Protocol optimisation

• None of these ^ apply to UC

• Lync Voice / Video Codecs already “self-optimised” and non-cacheable

• Can impact on remote users

• Previous known issues where Riverbeds have reset TCP traffic for STUN/MSTURN

Crypto Appliances

• Usually a mandated requirement for PSN or Official / Secret / Top Secret environments

• Crypto encryption adds latency to already encrypted media

• Business needs to accept the risk if a mandated requirement – i.e. NSA Suite B complianceWeb : northernucusergroup.co.uk - Twitter @northernucuserg

#5 “Blue Light” Emergency ServicesCertain scenarios may require an additional solution to a Lync deployment

• Self contained phone solutions that must have “dial tone” to designated handsets

• No dependency on underlying network or infrastructure services

• Police Command Centre

• Ambulance Control Room

• Nuclear Site License Requirements

• HSE obligations to edge sites that are either on a non-resilient LAN connection or have no local UPS

• Typically manufacturing environments

• Must be able to provide dial tone in the event of:

• Network outages (Planned and Unplanned)

• LAN / WAN / SIP Trunks

• Loss of Network or Power to none UPS protected edge sites

• Unforeseen circumstances


#5 “Blue Light” Emergency ServicesCertain scenarios may require an additional solution to a Lync deployment (cont)

• Sim-ringing mobiles will not be sufficient in these scenarios

• Lack of signal

• Mobile phones will not be allowed in Secret / Top Secret areas

• In the event of a Nuclear Emergency the cellular network will be taken down by the Police

Potential solutions could include

• Analogue gateways registered in Lync (network planned downtime needs to be carefully considered)

• Qualify length of cable runs required to support, anything over 5,000 meters will be a challenge!

• Understand limitations to supporting analogue devices in Lync

• Analogue Gateways connected to DEL’s with handsets connected directly to the Analogue Gateway

• Self contained PBX only providing analogue services

• Finding a resilient PBX that doesn’t require IP Services will be a challenge

• SBC’s connected to ISDN circuits in addition to SIP Trunks

• Still has a reliance on the underlying network and infrastructure being availableWeb : northernucusergroup.co.uk - Twitter @northernucuserg

#6 Sonus SBC’sTwo commonly deployed models as part of a Lync solution

• Sonus SBC1000

• 3 DSP’s

• Supports 160 concurrent transcoded SIP sessions

• Single PSU

• Sonus SBC2000

• 6 DSP’s

• Supports 600 concurrent transcoded SIP sessions

• Dual PSU’s

• Release Update specifics

• From release 3.2 and above

• Each DSP can support 200 concurrent sessions

• SBC’s can be shipped fully loaded and only need licenses to enable features

• DSP’s are not user serviceable parts, need to be preloaded prior to shipping

• Release 4.0 and above – Java is no longer required client-sideWeb : northernucusergroup.co.uk - Twitter @northernucuserg

#6 Sonus SBC’sDeployment specifics

• Use a naming convention that won’t confuse troubleshooting call flow issues

• If enabling MTLS ensure PTR records exist in the reverse lookup zone for the SBC and the Lync FE

• If deploying SBC’s without using the wizard and connecting to ISDN, configure Q.850 to SIP Override

• Default RTP/RTCP Port Pairs may need increasing

• IP call leg to TDM call leg = 600 calls, SIP call leg to SIP call leg = 300 calls

• To support more than 300 concurrent SIP to SIP calls the Number of Port Pairs will need increasing

• Configure SNMP on SBC’s and send SNMP traps into existing monitoring solutionsWeb : northernucusergroup.co.uk - Twitter @northernucuserg

#7 PBX Inter-Op / MigrationOption 1 (least preferred)

• Place SBC and Lync behind existing PSTN Presentation / Legacy PBX

• Typically connected to the legacy PBX by a QSIG connection or SIP Trunk

• Disadvantages

• Legacy PBX still has control of inbound / outbound routing

• Enabling of EV for Lync users requires ongoing PBX changes

• Lync project team will not be able to fully manage call control without assistance from the PBX engineer


PSTN

#7 PBX Inter-Op / MigrationOption 2 (preferred)

• Place SBC “in-line” between PSTN and legacy PBX

• Still connected to the legacy PBX by a QSIG connection or SIP Trunk

• SBC can cache msRTCSIP-line user attributes from Active Directory

• Inbound PSTN Call

• SBC performs lookup for msRTCSIP-line

• If the msRTCSIP-line is matched

• Inbound call is routed to Lync

• Lync performs a RNL for SIP Address

• Endpoint Rings

• If the msRTCSIP-line isn't matched

• Inbound call is routed to the legacy PBX

• Advantages

• Enables a controlled migration from legacy PBX to Lync

• SBC / Lync has complete call control for inbound / outbound PSTN calls


PSTN

#8 RCC / Client-Side Plug-In’sRCC – Remote Call Control

• Enables users enabled for Remote Call Control to use the Lync 2013 client to control a legacy PBX handset

• Only useful when a computer is tied to a desk

• Requires a Plus CAL

• Requires a CSTA Gateway eg. CUPS/AES (and relevant licenses)

• Deprecated in Skype for Business with the introduction of Call via Work

• CvW – Enterprise Voice for legacy PBX Users

• Leverages Direct SIP connectivity between legacy PBX and Skype for Business

• Users can make voice calls using any PSTN phone, including existing PBX endpoints

• Skype for Business dials out to the PSTN or a desk phone to connect the user, then connects with the far-end destination

• Presence update and call control from Skype for Business client


#8 RCC / Client-Side Plug-InsClient Side Plug-ins

• CUCILync / Avaya ACA

• Plug-in to Lync client controlling Cisco / Avaya PBX

• Allows the use of Lync for IM&P but uses a legacy PBX for call control

• Avaya ACA disables video within the Lync client by default

• Cisco recommends disabling everything apart from IM&P features within Lync Client

Both Solutions

• Vendors will state how good this solution is – It’s not!

• No EV configuration will apply, normalization will use company_phone_number_normalization_rules.txt

• Confusing user experience

• Unable to leverage Edge Services

• Unable to use Lync Audio/Video Conferencing

• Unable to use Remote User and Federation Audio/Video

• Additional software to manage / Additional training requirements / Additional licenses may be needed PBX side


#9 Service Visibility / MonitoringQoE

• Collects usage and QoE data for communication sessions

• Important Role – Should be deployed as part of any EV deployment

• Records information about each session

• Who called who, endpoints that were used etc.

• Uses MOS values to report on the perceived quality of the session

• MOS-LQ – Quality of audio for listening purposes only, doesn’t take into account delay, echo etc.

• MOS-CQ – Quality of audio in each direction, including bidirectional effects

• Provides MOS scores for:

• Listening MOS-LQ (stream played to user)

• Sending MOS-LQ (stream being sent to user)

• Network MOS-LQ (stream being played to user, including codec used, packet loss, jitter etc.)

• Conversational MOS-CQ (prediction of the narrowband audio stream being played to the user)


#9 Service Visibility / MonitoringKHI – Key Health Indicators

• Performance counters with thresholds aimed at revealing user experience issues

• Used in addition to standard Lync monitoring tools

• Creates a Excel based scorecard to determine server health within a Lync deployment

• Run included PowerShell script on each Front End Server

• Creates a performance monitor data collector and polls every 15 seconds by default

• After collection period complete the KHI Excel Spreadsheet

• Compare the results in the scorecard to the recommended thresholds


#9 Service Visibility / MonitoringKHI


#10 Exploit The Available ToolsMicrosoft Tools

• TechNet

• Lync Bandwidth Calculator

• Network Planning, Monitoring and Troubleshooting Guide

• Channel9 Deep Dive Videos

• MVA – Microsoft Virtual Academy

• Lync Rollout and Adoption Success Kit (RASK)

• Lync Planning Tool?

• Limited Use

Lync Community Tools (Remember to donate!)

• Ken Lasko’s Lync Dialing Rule Optimizer https://lyncoptimizer.com/

• Richard Brynteson’s Lync Validator https://lyncvalidator.com

• Pat Richard’s PowerShell Scripts https://ehloworld.com

• TwitterWeb : northernucusergroup.co.uk - Twitter @northernucuserg

https://lyncoptimizer.com/

https://lyncvalidator.com/

https://ehloworld.com/

Q&A

21Web : northernucusergroup.co.uk Twitter @northernucuserg

TenTips_GlenBrumpton_NUCUG_Feb_1.1_Cleaned

Documents

lync sue calls bob

pickup web

bob isnt

bob ill

pickup deployed sue

pool quorum

use of group

fe servers