Print Friendly
Comments

Check a Performance Counter

Test a variety of operating system and resource utilization metrics.

Task Parameters

Counter: The performance counter to be monitored. Current pre-defined counters are;

  • Memory Consumption
  • Memory Consumption (Virtual)
  • Processor Utilization
  • Processor Utilization (per-Core)
  • Logical Disk Utilization
  • Logical Disk Utilization (per-Disk)
  • Physical Disk Utilization
  • Physical Disk Utilization (per-Disk)
  • Battery Charge
  • System Up Time

Or use this field to specify any other available performance counter as described below.

Peak Sample (Utilization Counters): The highest values recorded over the test period will be considered (samples are otherwise averaged).

On Battery Only (Battery Charge): If mains power is online the counter will always return 100%. If mains power is offline the counter's value is capped at 95%.

User-Specified Counter

If the pre-defined counters don't serve a desired purpose you may instead specify any other available counter using the following syntax;

{Category}\{Counter}[?{Instance}|!{Instance}][*{Scale}][^%|^{Minimum}+{Increment}]

Category\Counter: The performance counter to be monitored (i.e. SQLServer:Buffer Manager\Page Life Expectancy). Required.

Include [?] OR Exclude [!] Instance: The name of a single counter instance which should be included (where prefixed with '?') as the sole instance to monitor or excluded (where prefixed with '!') as the only instance not to monitor. Generally this serves to configure the task to either monitor only the _Total instance or every instance other than _Total. Optional.

Multiply [*] BY Scale A positive integer used to scale all counter values up before comparison with an alert threshold. A typical use would be to convert a counter which returns fractions of a second into a whole number of milliseconds (*1000). Optional.

Threshold [^] AS Percentage [%] OR Minimum plus [+] Increment OR Maximum less [-] Decrement. Determines the 11 available alert thresholds. Default is ^0+10 (alert thresholds at; 0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100).

Most pre-defined counters have an underlying definition expressed using the same syntax;

  • Memory Consumption (Virtual): Paging File\% Usage?_Total^%
  • Processor Utilization: Processor\% Processor Time?_Total^%
  • Processor Utilization (per-Core): Processor\% Processor Time!_Total^%
  • Logical Disk Utilization: LogicalDisk\% Disk Time?_Total^%
  • Logical Disk Utilization (per-Disk): LogicalDisk\% Disk Time!_Total^%
  • Physical Disk Utilization: PhysicalDisk\% Disk Time?_Total^%
  • Physical Disk Utilization (per-Disk): PhysicalDisk\% Disk Time!_Total^%

Alert Conditions

Pre-defined counters return one or more numeric values in the range 0 to 100 which may be evaluated for an alert condition based on being less-than (or equal to) or more-than (or equal to) a percentage threshold between 5 and 95. The exception is the System Up Time counter which is instead compared to an interval of 1 to 12 hours, days, or weeks.

User-specified counters include the range and spacing of alert thresholds in the counter definition (or default to 0-100 where not specified).

Note that utilization counters require at least one interval to have elapsed before the utilization may be evaluated (over the task frequency).

7 September 2016