Skip to content
Rohit Macherla
GitHub
LinkedIn
StackExchange
RSS

Wiring it up - Integrating Salesforce with Prometheus

monitoring, salesforce2 min read

This is a part of series of blog posts about advanced monitoring techniques for your Salesforce.com application so your business can rely on the tech team.

So far, we've covered:

  1. A case for better monitoring
  2. Slack alerts for Salesforce.com
  3. Taking it to the next level for monitoring the Salesforce.com Enterprise
  4. Installing Prometheus on Heroku
  5. Prometheus metrics and API structure
  6. Exposing a Salesforce API that Prometheus can call

Back in Chapter 4, when we installed Prometheus on Heroku, we also setup a configuration file that it uses. It's the prometheus.yml file, which defines everything related to fetching data from other systems and recording and alerting rules to run.

This is the place we are interested in now.

Scrape Config

The main section of the config file is the scrape_configs. Each entry within this has information about server(s) to get metrics from. The act of getting this data is called scraping.

As we already know the URLs of the Salesforce API where our metrics API is hosted, we can use the static_config section and specify the targets. We can also configure how frequently Prometheus should poll this server via the scrape_interval available to be configured globally and also can be overridden at each scrape config level.

There is also config available to specify what authentication mechanism should be used. It supports Basic Auth, OAuth 2, or you can even specify an Authorization header as well. For our example, we've already exposed a Salesforce API that is publicly visible on a Site URL. When you are building this for your platform, please consider a good auth mechanism.

The realtime info about all these targets is available at the /targets endpoint of your Prometheus system.

Default Scrape Config

When we setup Prometheus, it comes with a default scrape_configs section that polls itself. That is, it's monitoring itself! Here's how it looks like:

1scrape_configs:
2 - job_name: 'prometheus'
3 scrape_interval: 5s
4 static_configs:
5 - targets: ['localhost']

Quite simply, it's calling itself to get metrics about its own server, every 5 seconds.

Adding Salesforce as a target

To add Salesforce as one of the servers to get metrics from, we need to create a scrape_config for it. Assuming that you've been following the tutorial, you'd have your Salesforce API setup from the previous chapter at /services/apexrest/api/v1/prometheus. Here's what we'll use for our example:

1- job_name: 'salesforce'
2 scrape_interval: 1m
3 metrics_path: '/services/apexrest/api/v1/prometheus'
4 scheme: https
5 static_configs:
6 - targets: ['<your site url here>']

Adding this to the existing Prometheus config file, the full file might look something like this:

1global:
2 scrape_interval: 10s
3
4scrape_configs:
5 - job_name: 'prometheus'
6 scrape_interval: 5s
7 static_configs:
8 - targets: ['localhost']
9 - job_name: 'salesforce'
10 scrape_interval: 1m
11 metrics_path: '/services/apexrest/api/v1/prometheus'
12 scheme: https
13 static_configs:
14 - targets: ['<your site url here>']

We've now configured Prometheus to call the Salesforce API that we've setup, which returns metrics in the format that Prometheus expects, and this happens once every minute.

So, we are retrieving all the Salesforce Org Limits into Prometheus.

Viewing the metrics in Prometheus

To confirm that we are getting these values, let us do a simple search on the Prometheus home page. I've deployed my app on Heroku following instructions from one of the previous chapters about Installing Prometheus on Heroku. We've prefixed each of these Limit metrics with an sfdc_limits, so a search on the home page brings them all out to us:

Prometheus showing Salesforce metrics on home page

Using a graph for sfdc_limits_dailyapirequests_current, we can see an increase in the number of daily API requests as I open my dev console and perform some SOQLs in there.

Prometheus showing SFDC Daily API requests for a period of 1 hour

That's it! We've now integrated Prometheus with Salesforce and can collect and alert metrics.

Next:

  • Wrapping the series!