Install and Configure Open Source Telegraf

Install and configure open source Telegraf to monitor your applications.

Telegraf uses input plugins (where the metrics come from) and output plugins (where the metrics go) in the configuration files. You can see all the supported plugins at Plugin Directory. Input and output plugins should be written in the Telegraf configuration file and configuration directory. The Telegraf configuration file (telegraf.conf) lists all available Telegraf plugins.

Telegraf uses the --config flag to specify the configuration file location and --config-directory flag to include files ending with .conf in the specified directory. On most Linux systems, the default locations are /etc/telegraf/telegraf.conf for the main configuration file and /etc/telegraf/telegraf.d for the directory of configuration files. For Windows platforms, it will be in the location where the telegraf zip is extracted.

For more information, see the Telegraf documentation, Get Started and Configuration Options. After you have downloaded and installed Telegraf, you are ready to begin collecting and sending data. To collect and send data Telegraf should be configured. Follow the steps below:

Procedure

Install open source Telegraf on the end point. If you have an instance installed, you can skip this step. To download and install a new instance of Telegraf, see the official documentation and search for the corresponding OS version from https://www.influxdata.com/time-series-platform/telegraf/ and https://portal.influxdata.com/downloads/.
On a Windows platform only, after downloading and extracting Telegraf files, besides telegraf.exe and telegraf.conf files, create a folder with the name telegraf.d that you use to run the helper script and monitor applications.
Run the helper script to configure Telegraf to send data to cloud proxy.
For more information, see Monitoring Applications Using Open Source Telegraf on a Linux Platform and Monitoring Applications Using Open Source Telegraf on a Windows Platform.

After you run the helper script, a cloudproxy-http.conf file is created and the output plugin is added to the file with the properties necessary for sending data and the input plugins are updated for OS metrics. For OS metrics configuration template, see Telegraf Configuration Details for Operating Systems. For information about the cloudproxy-http.conf configuration template, see Sample Configurations.
Using open source Telegraf, you can collect metrics from different types of application services.
- If an application service (curated plugin) that is supported by Application Monitoring in VMware Aria Operations is running on the end point and you want to monitor it, update the Telegraf configuration file or directory with necessary inputs for Telegraf.
  For a list of supported application services (curated plugins), see Supported Application Services. For the list of configurations, see Telegraf Configuration Details for Supported Application Services.
- For unsupported (non-curated plugins) application services, update the Telegraf configuration file or directory with the necessary inputs for Telegraf. For more information, see the Telegraf documentation Plugin Directory. For example, if you want to collect data for Ethernet device stats (device name is eth0) you can create the ethools.conf file in the Telegraf configuration directory and add the following content (for more information, see the documentation about Ethtool Input Plugin):
```
[[inputs.ethtool]]
  # List of interfaces to pull metrics for
   interface_include = ["eth0"]
```
- Whenever the telegraf-utils script is executed with Open Source Telegraf, the cloudproxy-http.conf file will be overwritten in the telegraf.d folder. You must maintain a different config file (for example, postgres.conf) for other custom input plugins to retain the configuration.
Start or restart Telegraf.