Windows Performance Counter task object in formatted JSON
{
    "Task_UID": "c93fd838f6e54093b82634c2ac2f4414",
    "CounterCategory": "memory",
    "CounterInstance": "",
    "CounterName": "qweqw4",
    "Machine": "Remote",
    "MachineAddress": "localhost",
    "Agent_UID": "5fc6e5ccf15a44df87c5bdb0cc63cb3e",
    "Aggregate": "Average",
    "MaxThreshold": 0,
    "MinThreshold": 0,
    "IgnoreErrors": "NO",
    "Device_Id": 7647,
    "Id": 16543,
    "Task_Type_Id": 23,
    "Name": "memory check"
}

Fields Description 

Field Type Required/Optional Description
Task_UID

String

optional

READ ONLY. Unique Task ID used by API to distribute gathered monitoring results.

Example:
"Task_UID""c7c6ffb9bae54df1b662fffb72964c77"
CounterCategory

String

required

A performance counter publishes performance data about an application. Categories include physical components (such as processors, disks, and memory) and system objects (such as processes and threads). System counters that are related to the same performance object are grouped into a category that indicates their common focus. For more information, please visit Microsoft official page .

Example:
"CounterCategory""memory"
CounterInstance

String

required

In some situations, categories are subdivided into instances, which track data about multiple occurrences of the object that a category relates to. Instances apply to the category as whole, rather than to individual counters. Every counter within a category has each instance defined for the category. For example, the Process category contains instances named Idle and System. Every counter within the Process category thus contains data for each instance, showing information about either idle processes or system processes. For more information, please visit Microsoft official page.

Example:
"CounterInstance"""
CounterName

String

required

End Counter name.

Example:
"CounterName""Available MBytes"
Machine

int32

required

Data Source can be located either on a local machine (where agent was installed), or at remote machine within LAN.

POSSIBLE VALUES:

  • Remote
  • Local
Example:
"Machine""Remote"
MachineAddress

in32

required

In case “Remote” Data Source was selected, address of the remote machine have to be specified.

Example:
"MachineAddress""192.168.1.56"
Agent_UID

String

required

 Agent UID.

Example:
"Agent_UID""5fc6e5ccf15a44df87c5bdb0cc63cb3e"
Aggregate

String

required

All received data will be aggregated on a regular basis, according to the adjusted device frequency.

POSSIBLE VALUES:

  • Maximum – the highest value from array will be taken
  • Average – value is calculated as an average of all intermediate values
  • Minimum – the lowest value from array will be taken
Example:
"Aggregate""Average"
MaxThreshold int32 optional Threshold. Exceeding will result alerting.

Example:
"MaxThreshold": 0
MinThreshold int32 optional Threshold. Exceeding will result alerting.

Example:
"MinThreshold": 0
IgnoreErrors

String

required

Each time during “Agent” <–>”Server” interaction Agent asks if there any new counters for Agent to check. In case there are  instructions  to gather stats on new counters Agents starts to gather them.

POSSIBLE VALUES:

  • NO – each failure in Counters polling will be reflected as error in reports.
  • YES – failures will be ignored.

Example:

"IgnoreErrors""NO"
Device_Id int32 required READ ONLY Field. Unique device id.

Example:
"Device_Id": 7647
Id int32

required

*Only for edit operations

READ ONLY Field. Unique task id.

Example:
"Id": 16532
Task_Type_Id int32 required The type of the task.

Example:
"Task_Type_Id": 23
Name String optional Task name.

Example:
"Name""anythingyouwant"

 

  • How MetricsView Works

    Performance counters are a built-in Windows mechanism that allows you to evaluate the performance and health of the system. Using performance counters, users and administrators can examine the operation of various components, including processors, drivers, databases, etc.

    In general, the process of monitoring Windows performance counters can be broken into three steps: collecting data, processing the received data, and notifying the user if the performance counters go beyond the specified limits.

    Collecting Data

    Windows collects system performance data, so-called Performance Counters, in real-time. The Performance Monitor tool is responsible for collecting these performance counters on Windows systems and can be used for manual system performance monitoring. It is a standard utility included in all modern Windows editions. The utility can be run either from the Start menu or from the command line or the search field by entering the perfmon command for Windows 8/10. After launching the utility, we see a standard panel in which we can add and remove counters, change the view and scale the data charts.

    To replace manual work with Performance Monitor and automate Windows Performance Counters Monitoring, MetricsView by Dotcom-Monitor is used. MetricsView setup includes three main steps –  setting up a MetricsView collector which enables the gathering of performance counter data from a target computer, installing the MetricsView software, called MetricsView Agent, on the target host and configuring the monitoring settings.

    The MetricsView Agent software is installed in a company network and capture the performance counters. Once installed and configured, the agent connects to the Dotcom-Monitor service and sends the performance counters statistics in a real-time mode to an assigned Dotcom-Monitor MetricsView Collector. The agent is able to monitor any performance counters available on the local host, as well as any additional performance counters of other machines on this network that the service has access to. To allow interactions with other remote services in the network you can allow the agent to log in under an account with the corresponding rights.

    The captured performance statistics includes all performance counters available for the system. The number of available performance counters varies depending on the version of the system.

    It is important to mention that the user under which the MetricsView Agent will run must have the necessary access rights.

    Processing Performance Counters Data

    MetricsView monitoring device can be set to monitor a specific performance counter in the selected collector with a specified frequency (starting with 15 mins and up to 1 min). All the data pulled from the collector is aggregated according to the device settings. You can set the device to aggregate the maximum, minimum, or average value of the monitored counter.

    Reporting and Alerting on Error Conditions

    In the device settings, you can configure maximum and minimum thresholds for the performance counter value. In the case when the value is beyond of set limits the system will generate an alert and send a notification to the provided notification address. This way you can be aware of your system bottlenecks as they occur.

    In addition, you can pull the MetricsView online report to review the statistics on the monitored performance counter. Use the online report to trace trends in your system performance and proactively react to its degradation.