You Asked, We Delivered: Introducing Windows Server Monitoring

Customers that use New Relic  to monitor their applications receive targeted, relevant information about performance bottlenecks affecting their apps. Whether it’s a decline in average response times, a change in throughput, or determining which web or database transactions are slow, New Relic makes it easy for developers to see where performance improvements can be made.

However, applications and services don’t run in a vacuum. They take up resources of one or more servers, whether running in a private datacenter or in the cloud. In a Windows environment, monitoring the health of these servers has typically required running ad-hoc tools or manually viewing Windows Performance Monitor output.

Since we introduced our Server Monitoring agent last year, we’ve received many requests to add support for Windows. We’ve promised that it’s coming and that day is here! We’re happy to announce that New Relic Server Monitoring now supports Windows servers. Beginning immediately, system administrators and developers can keep an eye on the health of any and all of their Windows server machines.

Whether you’re running a .NET app or an app written in another language – Windows Server Monitoring provides visibility into the performance and health of your Windows servers. In this post, I’ll review the features of Windows Server Monitoring and how it will benefit developers, system administrators and DevOps teams.

The 3,000 Foot View
After installing the Server Monitoring agent, server health metrics are automatically collected and sent to the New Relic SaaS service. The first place to get a glimpse of the state of your servers is the Servers Dashboard page. It shows the four key server metrics and a “traffic light” that indicates the overall health of the server.

Server Monitoring Dashboard

The Server Monitor Dashboard

In the example above, the qa-win2003-04 and qa-win2008-03 servers are reported as having disk space that is over 90% full.  Windows Server Monitoring uses default thresholds to indicate whether a specific metric is in a warning status (yellow) or a danger status (red). In the case of disk space, it uses 85% as the warning threshold and 95% as the danger threshold. The thresholds can be easily configured per metric (e.g., just for Disk I/O) and per server, or globally for all servers using the gear icon as shown below.

Threshold levels are easily configurable

Threshold levels are easily configurable

Normally, a system administrator wouldn’t want to continuously check a dashboard to see if a server is in trouble. Clicking the Configure server alerts link (seen at the bottom of the Threshold settings dialog) lets you configure email alerts that are sent whenever a metric threshold, such as Fullest disk % is breached.

Drilling Down on a Problem
With Windows Server Monitoring, you get alerts about problems such as high CPU and memory utilization, network spikes, and disk space issues before they effect your application.  That’s when you’ll want to drilldown further by going to the Overview page for a particular server as shown here.

Overview page for a specific server

Overview Page for a Specific Server

Just as our application agents allow you to view details of a web transaction such as a transaction trace or slow SQL query, Windows Server Monitoring gives you a detailed view into the health of the server over time. At a glance, you can see the trend in CPU usage, memory usage, including distinguishing between physical RAM and page file usage, and much more.

Let’s say you see that there is excessive memory usage on one of your servers and want to know which application is requesting so much of the system’s memory. On the Server Overview page (above) you can see a list of the top memory consumers by application.

Clicking the Processes heading link takes you to another detailed view, this time of the top memory and CPU consumers. The ease with which you can drill down for further detail is a hallmark of New Relic’s user interface.

The Processes Dashboard gives you a graphical view of the top memory and CPU consumers, and just like anywhere else in New Relic, you can expand the timescale to see historical trends or narrow the timescale to focus on a particular period of time.

Processes sorted by memory consumption

Processes sorted by memory consumption

Summary
As you can see, Windows Server Monitoring provides both a high level overview of server health as well as the ability to dive deep into the performance of the server infrastructure. While Server Monitoring is a great standalone FREE product, combining the power of New Relic’s Real User Monitoring, Application Monitoring, and Server Monitoring provides true end-to-end visibility of the entire application stack!

Try Windows Server Monitoring Today!
Windows Server Monitoring is available now and is included FREE of charge in all our subscription levels.
If you already have a New Relic account, click here to download and install the agent. If not, there’s never been a better time to try New Relic. Create your free New Relic account and get started with Windows Server Monitoring today! We think you’ll like it.

For more information, see our Windows Server Monitoring documentation.

Bob Uva is a Senior Software engineer working on New Relic's .Net Agent. View posts by .

Interested in writing for New Relic Blog? Send us a pitch!