nmon TA required?

[a212830]

Hi,

I installed the NMON app on my Linux search-head. I also ran the inventory search, but nothing came back. Is a TA required for this app? Is NMON itself required to installed separately?

[guilmxm · Accepted solution]

Hi,

Please review the Help page within the App (the question mark icon in Home page), you will find many useful information and also deployment scenarios)

Thus, to answer your question, this is totally dependent on what you to do, and can be easily customized to your needs.

1. Is nmon required to be installed separately ?

The App brings pre-packaged Nmon binaries for Linux x86 and x86_64 (currently Version 14g, generic flavor).
When you activate the nmon_helper, it will launches the pre-packaged binary only if nmon is not installed within your system.

If you can, I recommend to use the nmon binary that comes with the Linux distribution you are running. (all distributions brings nmon available in their repositories)

2. Is the TA required ?

Not necessary, depending on what you want to do.

You are running Splunk in cluster, if your goal is to:

• Get performance data of all members of your cluster: then yes the TA-nmon is required for non peer nodes (master, deployment servers and head search) to have the data collected from nodes and stored in the replicated index

• Get performance data from various hosts out of your cluster, then no you just need to install the PA-nmon on peer nodes (which is the bundle configuration + required elements to get perf data from peer nodes if wanted) and the normal App in head search.
  And finally deploying the TA-nmon to hosts that will stream their data to the cluster.

• Get Performance data from nmon files generated by a third party process and collected in centralized shares, then the normal App or the TA-nmon (through heavy and light forwarder) can be used to watch for external directories and manage nmon data when required

Some more information:

• An standalone Indexer instance can simply use the normal App to collect the Nmon data, upon installation you just to need to activate the local data collect (or after restart in Settings page)

• Peer nodes of a cluster will use the PA-nmon app to collect and index Nmon data in the distributed index
  The PA-nmon also contains the bundle configuration for the clustered nmon index, and this should be pushed by the master node

• Heavy Forwarder / Light Forwarder can run the TA-nmon app to collect Nmon data (local or central shares) and stream it to an indexer instance (standalone instance or cluster)

As you are in a cluster environment, you can review the Cluster deployment scenario (scenario 2) which is a proposal configuration.

The normal App needs in deed to be installed in search head, but a search head does not index data himself, that's why the scenario proposes the installation of a Light Forwarder running TA-nmon for all non peer nodes of a cluster you want to manage with Nmon

Please let me know if it answers to your questions, and contact me if you can't get it to work as you whish.

Guilhem

[krishananth]

Hello,

I would like to collect and forward nmon data from different unix machines (having light Universal forwarder installed) to a standalone instance (splunk receiver). It looks like I need to use TA-nmon. i.e. install within %splunk-forwarder%/etc/apps in my unix machine.

Im unable to find the download location of TA-nmon unix installation file.

Could you please help me on this?

• Ananth

[guilmxm]

Hi,

Please think in opening new questions instead of commenting existing and answered, to improve your question visibility.

The TA-nmon package is located within the "resources" directory of the main package you download in Splunk App.
Do not hesitate to read the inline documentation to help you in your deployment.

Thank you for using the App 🙂

Guilhem

[a212830]

Awesome. Thanks.