Topic Last Modified: 2012-10-22

Between various websites, Microsoft Knowledge Base articles, and Lync Server Resource Kit tools, administrators who encounter problems when running Lync Server are never far from a way to solve those problems.

Obviously there is no way to guarantee that you will never encounter problems with Lync Server 2013 because Lync Server can be affected by many things—like network crashes and hardware failures—that the product itself cannot control. By implementing health monitoring, administrators can identify potential problems before they turn into actual problems. For example, administrators can use Lync Server monitoring to identify trends and tendencies. For example, a steady increase in the number of audio/video conferences might suggest a need to add capacity before the system becomes overloaded.

In a similar fashion, administrators can use System Center Operations Manager to do such things as issue real-time alerts when specified events occur, and to run synthetic transactions that proactively test the system. Synthetic transactions are used in Lync Server to verify that users are able to successfully complete common tasks such as logging on to the system, exchanging instant messages, or making calls to a phone located on the public switched telephone network (PSTN). For example, periodically running these tests can alert you to potential problems with users logging on to Lync Server, and give you a chance to correct the problem before your support team is flooded with calls from users unable to make a connection. By using System Center Operations Manager to run these synthetic transactions, administrators can routinely monitor their deployment of Lync Server continuously for 24 hours every day without having to do much of anything beyond responding to any alerts that might be issued.

For Lync Server 2013, the Management Pack for System Center Operations Manager is also able to detect "external" issues that can adversely affect Lync Server. For example, administrators can be notified if Internet Information Services (IIS) goes offline, system resources on a Lync Server computer fall below a specified amount, or a Lync Server computer experiences a hardware failure.

Health configuration in Lync Server 2013 is built around System Center Operations Manager and the use of Lync Server Management Packs. These Management Packs include a number of new features and enhancements, including:

The Management Packs also include a variety of features to help detect and diagnose provide real-time visibility into the health your Lync Server deployment. These features are listed in the following table.

Management Pack Features

Feature Description

Synthetic Transactions

Windows PowerShell cmdlets that can be run from various locations to ensure that end user scenarios such as sign-in, presence, IM, and conferencing are readily available to end users.

Call Reliability Alerts

Database queries for Call Detail Records (CDR). These records are written by Front End Servers to reflect whether end users were able to connect to a call or why a call was terminated. These queries result in alerts that indicate when a wide range of end users are experiencing connectivity issues for peer-to-peer calls or basic conferencing functionality.

Media Quality Alerts

Database queries that look at Quality of Experience (QoE) reports published by clients at the end of each call. These queries result in alerts that pinpoint scenarios where users are likely to be experiencing poor media quality during calls and conferences. The data is built upon key metrics such as packet latency and loss, metrics that are known to directly contribute to call quality.

Component Health

Individual server components raise alerts by using event logs and performance counters. These alerts indicate failure conditions that can severely impact one or more end user scenarios. These alerts can also indicate a variety of other failure conditions, including services not running, high failure rates, high message latency, or connectivity issues.

Dependency Health

Failures can occur for a variety of external reasons. The management packs now monitor and collect data for some of the critical external dependencies that might indicate severe issues, including IIS availability, CPU and memory usage of servers and processes, and disk metrics.

The alerts issued by the system have been classified into three general categories:

In This Section