{"id":531,"date":"2015-06-29T08:51:00","date_gmt":"2015-06-29T08:51:00","guid":{"rendered":"https:\/\/blogs.msdn.microsoft.com\/premier_developer\/2015\/06\/29\/web-app-performance-are-you-listening-to-your-logs\/"},"modified":"2019-03-07T13:36:27","modified_gmt":"2019-03-07T20:36:27","slug":"web-app-performance-are-you-listening-to-your-logs","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/premier-developer\/web-app-performance-are-you-listening-to-your-logs\/","title":{"rendered":"Web App Performance : Are you listening to your logs?"},"content":{"rendered":"

In this post, Principal Application Developer Manager, Reed Robison <\/strong>explains how a periodic review of IIS Log data, often ignored and archived, provides valuable insight into the operational health and performance of web applications.<\/p>\n

\n

 <\/p>\n

We recently posted about how a focus on DevOps<\/a> is changing the way IT Operations and Developers work together to deliver higher quality solutions. IT\/Dev collaboration takes practice and finding ways to effectively work together might be a little different in every organization. Examples are always helpful, so let\u2019s talk about web application performance.<\/p>\n

Performance Lifecycle<\/h3>\n

It is pretty common for teams to do performance testing before releasing an application. While testing, developers intensely analyze and optimize code. After applications are released into production, they rarely receive the same level of operational scrutiny. The focus tends to shift toward basic monitoring of a few high level metrics like CPU, memory, or response time. While this may be sufficient for daily monitoring, it can mask details and trends that are<\/em> captured in logs and creep in over the lifecycle of an application. If testing is not a part of every release cycle, it\u2019s not uncommon to for new bottlenecks to manifest through code updates, changes in usage patterns, or a steady increase in load as business grows. These little problems may go unnoticed until they become big problems.\u00a0 A good way to regularly and comprehensively assess operational environments is through RAP as a Service<\/a>, however, you have data available right now that provides operational insight into the health of your web apps.\u00a0 Almost every organization records and archives web logs.\u00a0 In this post, we\u2019ll look at some tips to use IIS Logs to quickly produce operational insights about the performance of your application.<\/p>\n

Using data you already have<\/h3>\n

\"\"In Premier Support, customers routinely engage us with performance and stability problems of web applications. While Testing Services<\/a> engagements can comprehensively help customers plan for and avoid capacity issues, not every application gets this level of attention. Similarly, there are many powerful diagnostics and monitoring tools that can help isolate problems after integrated into a solution, but they can require significant code and architectural changes. Sometimes, we have to work with whatever data we can get out of a live production system, and very often that means IIS Logs. Most operational teams can provide a history of IIS logs for any web application. It turns out that IIS Logs are generally full of valuable trend data that developers can use to isolate performance problems in their code.\u00a0 They also provide a historical record of valuable data to analyze trends over a period of days or months.<\/p>\n

In order to size up the health of a web application very quickly and with limited data, consider three things:<\/p>\n