Monitoring multi-instance counters in NetCrunch
Learn how to easily create counters with many instances in NetCrunch. Managing a long list of alert conditions for a node might be complicated when it's done "one by one". NetCrunch allows you to manage alerting & reporting settings in a straightforward way by using Monitoring Packs.
What is a Monitoring Pack?
A monitoring pack is set of rules that determines what is monitored on a particular node and what conditions have to be met for NetCrunch to trigger an alert.
There are two types of Monitoring Packs
- Static - Users can manually selecting which nodes this set of rules should apply to.
- Automatic - Users can automatically set filtering criteria for which nodes this set of rules should apply to. This article will focus on that functionality.
Creating an Automatic Monitoring Pack.
'Discarded outbound Packets' is a Windows performance counter with multiple instances that we will use in this example of how to add multiple alert conditions at once.
To create a new monitoring pack in NetCrunch, open Alerting & Reporting Settings: ToolsAlerting & ReportingSettings and click New Monitoring Pack. In the newly opened window, choose Automatic Monitoring Pack
Configuring Automatic Monitoring Packs is simple. The first thing to do is enter the name of the Monitoring Pack in the top left corner of the screen. You can now set restrictions. In this case we will set the restriction to Windows Only, as this counter will be tracked on Windows systems only.
Next, switch to the Alerting & Reporting tab below and click Add Alert in the Add Monitoring Event window. Double-click New Event for Windows Performance Counter.
In the Alerting Rule window we need to select the counter and its threshold value. To do that, click select counter. Now search for the object that we want to monitor. Set Performance object to Network Interface, select Packet Outbound Discarded and click on All instances on the right to make sure that all instances on all machines will be monitored.
After clicking Add, the counter is set. The last thing to do is to set the actual threshold value. in our example we will set it to: " > 10"
Please take note that threshold configuration may vary depending on network conditions and overall network performance characteristics.
A monitoring pack can consist of multiple counters and alerting conditions, but in our example we have created one with a single alerting rule.
The Node Filtering tab is where you set what nodes this monitoring pack should be applied to. This is an automatic & simple way of applying it to even large numbers of nodes in one step. In our case it is a simple filtering condition. This monitoring pack will be assigned to all Windows Server nodes.
Click Add Condition and set Operating System is equal to Windows Server.
Overall configuration of the filtering should look like this:
After Saving the new Monitoring Pack, NetCrunch will now monitor these counters, and generate alerts when any interface exceeds the defined threshold.
Creating a Top Charts view based on a Monitoring Pack
The best way of displaying the values of counters that were added in this monitoring pack is to create a Top Charts view for them.
Click the Network Atlas view in the Atlas Tree and select Dashboards on the right, then click Top Charts below.
Click the + sign on the far right of the screen, hover over the Top Chart and select New Chart.
Click [...] under the counter, use search to find the counter, type "Outbound", and extend "Network Interface". The Top Chart may include a particular interface, but in our example we are selecting Packets Outbound Discacrded
This is a basic configuration of the Top Chart view. NetCrunch will now display the "Most Problematic" nodes on the top of the list. In this way you can easily identify servers with the highest level of discarded outbound packets.
