Lync was released almost two years ago, and adoption is growing. Although Microsoft has delivered numerous improvements throughout the years, it is still very challenging for administrators to ensure a good quality of service for Lync, especially for audio/video calls.
You will find here some of the most common issues encountered for Lync 2013 with tricks that can help you troubleshoot and solve them.
Issue #1 Virtualization
For example, one of the common issues that can be encountered comes from a misconfiguration of the VM host. The team in charge of the virtualization should work very closely with the Lync team and follow the guidelines that Microsoft has released in a whitepaper.
Issue #2 Quality of Service (QoS)
If you find a considerable amount of packet loss and you cannot increase the bandwidth, it is highly recommended to implement Quality of Service (QoS). The QoS allows you to assign the network packets for audio and video communication higher priorities so that communications can be faster and more reliable.
Issue #3 Front-End and Backend Performance
It is critical to pay close attention to Key Performance Indicators on both the backend and the front-end servers in Lync 2013. The SQL team should treat Lync as a very specific real-time application. The real-time side of it means that the accesses to the database needs to be very quick. The Stored Procedures have to be executed in a few milliseconds at the most. Therefore, you must pay close attention to CPU, Memory and Disk I/Os. Similar issues happen on the front-end servers, as they host a SQL Express instance the disk performance needs to be very high, that is new to 2013.
What to do to control performance?
Keeping Lync 2013 service available and performing well requires that you keep track of multiple values including performance metrics, quality of service delivered to end-users, and being able to look at the evolution of the performance as well. Within medium to large organizations, it is also critical to establish SLAs and KPIs to report on, in order to ensure no decrease of performance over time. It is important to keep in mind that every Lync deployment can experience issues at some point, and especially user complaints as the number of users will grow over time, and that end-users will make calls from remote locations with poor network bandwidth.
How GSX Solutions can help?
In order to help administrators in this monitoring, GSX Solutions has released a new version supporting Lync monitoring. Different things were targeted in this version:
- Performance counters on the front-end and backend servers.
- Synthetic transactions for testing service delivered to end-users
- Reporting on key performance indicators and preconfigured reports for tracking SLA and KPI.
GSX Solutions has gathered about 50 critical performance counters out of the 8000 that Lync provides and simulates 14 end-user transactions targeting features like user authentication, presence, instant messaging or video calls, just to name a few.
These tests are really performing a complete set of transactions that occur while a user consumes Lync services. For example, here are the detailed steps involved in the scenario testing the instant messaging feature:
- Registration of test user 1 and test user
- Authenticate test user 1 and test user 2
- User 1 creates invitation for peer to peer instant message
- Send peer to peer instant message to user 2
- Verify that instant message was received by user 2
- Issue a “bye” peer to peer instant message by user 1
- Unregister user 1 and user 2
GSX Solutions can alert and report on the availability and execution time of all of these tests too, so that you can keep control over your Lync environment. Find out more about GSX Monitor & Analyzer for Microsoft Lync:
Olivier Raynaut is a GSX specialist for Microsoft Platforms