Before we start working, we should also look at the tools. What tools are there and for what use are they intended? The following list gives you an overview of the currently available offer:

• Azure management portal: The Azure portals provide dashboards that display specific metrics for Azure services that might not be available through APM tools or profilers. You can view, for example, the average latency of Azure Storage, or the rate for messaging within an Azure IoT Hub.
• System specific tools: Some services (including third-party services) frequently provide their own monitoring and analysis utilities. Examples include the SQL management views, and the Query Performance page in the SQL Management Studio. These tools are dependent on the nature of the service being provided. For example, the Query Performance page in the SQL Management Studio lets you view the details of the query execution plan for recently executed SQL statements. If you have an understanding of how SQL Server optimizes queries, you might be able to use this information to make your queries more efficient.
• Performance counters: Performance counters refers to the measured values which are captured by the Windows operating system. A large number of different counters are available. However, many of them only record data at a low level, such as the counter for speed at which drives are read or written, the counter for memory available on the computer, or the counter for the CPU load of an application. These metrics are gathered automatically and the collection process is transparent to applications and services. Services can also create their own custom counters. For example, ASP.NET adds counters that track the request queue length and the rate at which requests are sent and responses received. Tools such as Windows Performance Monitor can be used to visualize performance counters. Applications and services may use them to trigger alerts if they stray outside an expected range. However, the performance counters only use sampling, so it is possible that one or more significant outliers could be missed.

• Log parser: Applications and services can generate their own custom diagnostic trace information and record it to log files (for examples the IIS Log, the Windows Event Log, or custom files created by applications for logging purposes). The data is typically text-based, but the format of the data might vary significantly from file to file. A log parser lets you quickly read and process the contents of different log files. You can also use it to perform analyses, for example, sorting and aggregating log records and filters. The protocol parser uses SQL-like syntax to specify queries about log data.

• Profilers: These tools provide a granular view of how an individual executable is consuming resources, enabling you to perform in-depth investigations into CPU, memory, disk, network, garbage collection, and threading. Most of these tools are based on Event Tracing for Windows (ETW), so they can be used to track and analyze code running on any version of Windows. A common feature of these tools is the ability to examine crash and memory dumps to help determine the cause of exceptions. Examples include ANTS Profiler and Visual Studio Profiler (for general application profiling), PerfView (for tracking CPU and memory-related issues), CLR Profiler (for tracking managed memory use), and Concurrency Visualizer.

• APM tools: APM tools are tools that can be used to monitor cloud applications. The capabilities of these tools are not limited to development and testing scenarios but are also used in monitoring production environments. APM tools are designed to minimize the administrative burden of using intelligent strategies, such as the automatic detection of dependent services. APM tools typically use monitoring techniques that do not involve the insertion of code probes. In fact, APM tools rely on system or runtime-provided trace capabilities and autonomous monitoring agents to capture operational data. Examples include Azure Application Insights and the third-party offerings from AppDynamics and New Relic.

I think that's enough for now. Detailed information on these and other tools can be found in the next chapter.