Servers for Windows monitor metrics

For the latest, most advanced host monitoring, check out New Relic Infrastructure.

Servers for Windows samples data every 20 seconds, and then communicates the data to the New Relic collector once per minute. This document summarizes the metrics collected by Servers for Windows.

Environment

Servers for Windows collects the following information about your hardware and software environment:

  • Server name
  • Number of CPUs (including vendor, model, speed, and cache size)
  • Amount of physical RAM available
  • Windows version name (for example, Windows Server 2003)
  • Windows Service Pack version
  • Windows version number
  • Processor architecture (x86 or x64)
  • Windows server monitor agent version

To view your environment data: From the New Relic menu bar, select Servers > (selected server) > Overview.

Memory

Servers for Windows collects the following memory usage metrics:

  • Used: Amount of physical RAM being used.
  • Swap: The percentage of page file space that would be used if running processes were paged to disk. This is measured in committed bytes, and does not represent actual swapping of pages to the pagefile. To view only current memory usage, disable the Swap metric by selecting its label in the chart.

To view metrics from the UI:

  • Memory metrics: From the New Relic menu bar, select Servers > (selected server) > Overview.
  • Processes by the amount of memory they use: From the New Relic menu bar, select Servers > (selected server) > Processes, and then sort by Memory.

Network

To view network data from the UI:

  • Snapshot of network I/O: From the New Relic menu bar, select Servers > (selected server) > Overview.
  • Detailed network statistics: From the New Relic menu bar, select Servers > (selected server) > Network.

Servers for Windows captures the following network metrics and statistics for each network interface card:

Network metrics
  • Network interface card name (for example, Local Area Connection)
  • Number of bytes transmitted per second
  • Number of bytes received per second
  • Number of packets transmitted per second
  • Number of packets received per second
  • Number of errors transmitted per second
  • Number of errors received per second
Network statistics
Network statistic meanings:
  • Bandwidth is the amount of data transferred per unit time. Tx/transmitted and Rx/received are totaled in the Bandwidth graph.
  • Packets per second is how many separate network packets have been sent and received.
  • Errors are simply the number of errors recorded for the interface by the OS.

File system

Servers for Windows measures the number of bytes used. The New Relic Servers product uses this data to estimate the rate at which your disk free space is filling up. To view disk utilization: From the New Relic menu bar, select Servers > (selected server) > Disks.

Disk I/O

The Disk I/O utilization measures the amount of time that a device is busy vs. idle. It is not based on throughput.

Servers for Windows collects the following I/O metrics for each disk:

  • Disk drive identifier (for example, C:)
  • Utilization: Read and write percentages
  • Amount of data written per second
  • Amount of data read per second
  • Number of writes per second
  • Number of reads per second

To view disk I/O data from the UI:

  • Snapshot of disk I/O: From the New Relic menu bar, select Servers > (selected server) > Overview.
  • Detailed disk I/O statistics: From the New Relic menu bar, select Servers > (selected server) > Disks.

CPU metrics

Servers for Windows measures the following CPU metrics:

  • Percent usage by system (kernel)
  • Percent usage by user processes

To view CPU data from the UI:

  • Snapshot of CPU usage metrics: From the New Relic menu bar, select Servers > (selected server) > Overview.
  • Processes by the amount of CPU they consume: From the New Relic menu bar, select Servers > (selected server) > Processes, and then sort by CPU.

CPU states

CPU states have the following meanings:

CPU states Explanation
I/O Wait Time that the CPU is idle and there is at least one input or output operation in progress.
Stolen

CPU time "stolen" from this virtual machine by the hypervisor for other tasks, such as running another virtual machine.

Your New Relic Servers charts will show increased Stolen activity only when the app has activity. For example, if resources are stolen but the virtual machine is not actively processing, no load shows. However, if resources are being stolen and the virtual machine is active (even to a low degree), the load spikes proportionally. The more stolen resources there are, the less activity the virtual machine needs to do to get a high load rating.

System Time used by the kernel and its associated processes. These include rebuilding RAID, handling network transmission, processing checksums, and other system "housekeeeping."
User Time the CPU has spent running users' processes.
Idle Anything between the top of your usage data and 100% (in white) is time when the CPU is not doing anything.

Processes

For each sampling interval, Servers for Windows reports the following metrics about your running processes:

  • The Account identity under which the process is running (for example, SYSTEM or a specific user account)
  • Process name
  • How many instances of each process are running (for example, two Notepad instances)
  • Percentage of CPU usage
  • Memory working set size

To view these metrics: From the New Relic menu bar, select Servers > (selected server) > Processes.

Access server monitor metrics via API

To query server monitoring performance data, you can use the New Relic REST API and the API Explorer. For example:

For more information, see the REST API's server documentation.

For more help

Additional documentation resources include:

Discuss New Relic Servers in the New Relic Online Technical Community! Troubleshoot and ask questions, or discuss Servers for Linux or Servers for Windows in detail.

If you need additional help, get support at support.newrelic.com.