Need help on CPU/Memory Performance Report

[bsivavani]

Hi,

We are planning to send CPU/Memory performance report (One month/Week data) on all windows servers to users.

Can you advice if there is any way to get this report ?

[benjaminsmith]

Hi @bsivavani,

From the Nagios XI home page, go to Details > Metric. You can view and download graphs for CPU / Memory by hostgroup. Below is a link to a short video on accessing performance graphs in Nagios XI.

Nagios XI - Performance Graphs & Graph Explorer

Let me know if you have any other questions.

[bsivavani]

Thanks for the update.

We would like to see Physical Memory service which we have created in Metric drop down.

Kindly let us know how to add custom service in Metric downtime ?

[benjaminsmith]

Hi @bsivavani,

If you're are using the NCPA agent, we can try to modify the syntax of the check command to make it available from the metrics component.

Otherwise, another option is to create a custom dashboard of the Physical Memory graph for the windows servers. This will create a customized view of those services which can be downloaded or scheduled in a report.

Understanding and Using Dashboards

[bsivavani]

I think second option will not work here because we have 500 servers and we need to go to each server physical memory graph and add it to dashboard manually.

First option may work using NCPA agent.
Can you provide more details on how to configure NCPA agent to get the Physical Memory under Metrics component ?

[benjaminsmith]

Hi @bsivavani,

You can use the NCPA configuration wizard to create the CPU / Memory checks. Once you create those checks, you'll need to review/edit the check commands to make sure they have the correct syntax. Go to Configure > CCCM > Monitoring > Host or Server > Common Settings to review the check commands.

For CPU, the -M arg should be cup/percent:

Code: Select all

check_ncpa.py -H host-address -t 'your-token' -P 5693 -M cpu/percent -w 20 -c 40 -q 'aggregate=avg'

For Memory, the -M arg should be virtual/memory/percent:

Code: Select all

check_ncpa.py  -H host-address -t 'your-token' -P 5693 -M memory/virtual/percent  -w 50 -c 80

Let me know if you're able to get it working.

[bsivavani]

We are using check_nrpe to get the Physical Memory on windows servers.
Is it possible to get physical memory under metrics component?

[benjaminsmith]

Hi @bsivavani

We are using check_nrpe to get the Physical Memory on windows servers.
Is it possible to get physical memory under metrics component?

Currently, the Metrics Component will not work when using check_nrpe for Windows servers, so using NCPA, if possible, would be the best option.

[bsivavani]

Hi,

NCPA option also rolled out.

Could you please let us know if there is any way (mysql query) to get the Last 1 week CPU performance data from MySQL database.

Kindly advice if there is any mysql query to get the raw data from Nagios server MySQL prompt ?

[benjaminsmith]

Hi @bsivavani,

The performance data is stored in RRD (round robin database) files. If you go to Home > Graphs > Graph Explorer you'll see a series to tabs to generate performance graphs by host, service, data type ..etc. Also, from the 'hamburger' menu in the upper right of the graph you have the option to download the image or the data itself.

Let me know if that's what you need.

Here's an overview of how the process works:
Nagios XI Performance Data Architecture