Installing and Managing Wavefront Proxies

In most cases, a Wavefront proxy must be running in your installation before metrics begin streaming to Wavefront from a host or application.

You can install a proxy as part of a built-in integration. Even collector agent integrations, such as the Telegraf integration, install a proxy. But for some custom integrations, you install the proxy yourself. This topic describes methods for installing Wavefront proxies and for managing proxy services.

If you don’t use the Wavefront UI to install the proxy, the installation procedures might require a Wavefront API URL <wavefront_api_url> in the format https://<wavefront_instance>.wavefront.com/api/ and an API token. To get an API token, see Generating an API Token.

Note Some installation procedures automatically configure these properties for you.

Proxy Host Requirements

• Internet access - run timeout 3s curl -fIsS <wavefront_api_url> from the host and make sure you get a response and not a timeout.
• Memory - you don’t need a dedicated host for running the Wavefront proxy. The proxy does not use a lot of CPU, memory, or storage. However, we recommend running the proxy on a host with at least 4GB of free memory.
• Operating system
  • Linux
    • Ubuntu 14.04, 16.04, 18.04
    • CentOS 6.5, 7
    • RHEL 6, 7
    • Debian 7, 8, 9, 10
    • Amazon Linux
  • Mac - MacOS Sierra (10.12)
  • Windows - Windows 8 and later

You can also run a proxy in a Docker or Kubernetes container.

Proxy Installation

If we don’t install a proxy for you as part of integration setup, you can install a proxy explicitly. You can script installation on Linux and Mac OS. You can also run a proxy in a Docker container.

Note: You can use a single proxy to pass input from many collector agents to the Wavefront service. In production, consider using two proxies behind a load balancer. See Proxy Deployment Options.

Installing a Proxy on a Single Host

To install and run a proxy on a Linux, Mac, or Windows host, or in a Docker container on a host:

1. Open the Wavefront application UI.
2. Select Browse > Proxies.
3. Select Add > New Proxy at the top of the filter bar.
4. Click the [Linux | Mac | Windows | Docker ] tab.
5. (Windows Only) Download the proxy.

6. Copy the script and run it on your host.

   Note On Windows, do not run the installer .exe file. Run the script instead.

7. After the proxy contacts the Wavefront service, the proxy name displays under “Checking for new proxies…” and the button label changes to Done.
8. Click Done and verify that your proxy is listed on the Proxies page. If not, follow the steps in Managing Proxy Services to start the proxy is running.

If you want to run a proxy in a Kubernetes container, log in to Wavefront, click Integrations, click Kubernetes, click the Setup tab, and follow the instructions to deploy a Wavefront proxy in Kubernetes, create a proxy service, and deploy Heapster.

Scripted Proxy Installation

In Linux hosts, you can use the Wavefront CLI to install the Wavefront proxy and to perform certain management tasks.

In Mac and Linux hosts, you can select the integration for the host and run only the command that installs the proxy, not the command that installs the Telegraf agent.

Managing Proxy Services

After installing a proxy, you can start and stop the proxy service, check service status, and view the logs that are generated by the service. See Logging for log configuration options.

Starting and Stopping a Proxy

You can start and stop a proxy by running the following commands on the proxy host:

• Linux

  $ service wavefront-proxy [start | stop | restart]
  

• Mac

  $ brew services [start | stop | restart] wfproxy
  

• Docker

  $ docker [start | stop ] <proxy_container_id>
  

• Windows

  $ cd C:\Program Files (x86)\Wavefront\bin
  $ ./nssm.exe [start | stop] WavefrontProxy
  

Checking Proxy Service Status

To check if the proxy is running, run the following commands on the proxy host:

• Linux

  $ service wavefront-proxy status
  

  You can view the proxy log at /var/log/wavefront/wavefront.log.

• Mac

  $ brew services list
  

  You can view the proxy log at /usr/local/var/log/wavefront/wavefront.log.

• Windows

  $ cd C:\Program Files (x86)\Wavefront\bin
  $ ./nssm.exe status WavefrontProxy
  

  You can view the proxy log at Program Files (x86)\Wavefront\wavefront.log.

• Docker

  $ docker ps
  

  To view the proxy log, run docker logs <proxy_container_id>.

Testing a Proxy

You can test that a proxy is receiving and sending data by sending it a JSON payload as follows:

1. In the proxy configuration file, set the JSON listener port to 3878.

   jsonListenerPorts=3878
   

2. Run the following command:

   curl --header "Content-type: application/json" \
   'http://<wavefront_proxy_address>:3878/?h=test_host&d=now' \
   -d '{"test.metric":{"value":1,"tags":{"key1":"v1","key2":"v2"}}}'
   

   where <wavefront_proxy_address> is the address of your Wavefront proxy.

3. In the Wavefront UI, select Browse > Metrics.
4. In the Metrics field, type test.metric.
5. Click test.metric to display a chart of the metric.

Upgrading a Proxy

Wavefront releases new proxy versions with new features periodically. See Proxy Release Notes

To upgrade the environment, you can select Browse > Proxies > Add New Proxy. If an older version of the proxy exists, this process replaces it.

Note On Windows systems, you might have to uninstall the existing proxy first. Use the uninstall process for Wavefront Proxy on the Windows system you’re using.

For Linux and Mac OS, can also upgrade a proxy from the command line as follows:

Proxy Troubleshooting