PromVIEW

PromVIEW is a Prometheus client library for LabVIEW applications.

This library allows you to instrument your LabVIEW code so you can expose metrics scraped by Prometheus. For more information, see Prometheus website :

Getting Started

Prerequisites

LabVIEW 2020 (Full, Pro or Community) or later
- The NI Web Server service must be enabled, configured to allow HTTP and with CORS enabled on the Prometheus port (default 9090)
A reachable and configurable Prometheus server

First Example

Install the PromVIEW package.
Open the LabVIEW project you need to instrument, or create one.
Add the Initialize Default Registry VI to your code (preferably in the init/setup section of your application)
Add a new web service to your project. Give it a name without space, for instance "promview-tester"
Create a new web resource VI. In the block diagram, add the GET metrics route VI from the PromVIEW palette.
In the web service properties, copy the URL of the above resource. For instance http://localhost:8020/promtest
Run your application.
Right-click your web service and click Start.
Use your usual web browser and go to the above URL. You should see page of plain text with default metrics values.

Connecting to Prometheus

Now that your code is instrumented, you can add the newly created web service as a new target to be scraped by Prometheus. To do so, edit the prometheus.yml file for your Prometheus instance and add a new target. For instance:

- job_name: 'PromVIEW Test' scrape_interval: 2s static_configs:

- targets: ['<Machine running LabVIEW>:8020'] metrics_path: '/promtest'

(Please make sure the indentation of this config is correct when pasting to prometheus.yml)

Then restart your Prometheus service (on a Linux host, service prometheus restart).

You can check if the target was successfully added to Prometheus by going to your Prometheus page ([http://:9090](http://:9090)), under Status > Targets.

If your target is up, then you should be able to query the metrics exposed (from the Graph menu on the Prometheus web page).

Instrumenting your LabVIEW Application

Metrics

PromVIEW provides the metric types that Prometheus is able to scrape. Simply create a new metric by using the Create VI for the metric type you need. By default, the metric is automatically added to the default registry. All there is to do next is to use the metric functions (like increment, set or observe) accordingly in your application.

Collectors

A collector contains a set of metrics that can be scraped together or in a different way from individual metrics. You can create custom collectors as per your application needs, and register these collectors to a given registry. You can also customize the way the collector will collect the metrics it owns.

To create a custom collector, just create a new class that inherits from Default Collector.lvclass.

Registries

A registry is a container for metrics and collectors. A default registry named _default_registry is created when you invoke the Initialize Default Registry function. You can create more registries by using this same function and passing a different name.

Registries in PromVIEW are session-based, meaning you may access a registry by using the Obtain Session function anywhere in your application.

A metric or collector may be registered to multiple registries.

You can expose metrics from a specific registry by adding a query parameter to your web resource VI. Please refer to the LabVIEW documentation if you are unsure of how to accomplish this.

pfoneill/PromVIEW