Skip to main content

Test Promtail Pipeline Stages Locally on Your Machine


If you use Loki as your log aggregation system, then you're likely familiar with Promtail, the agent that ships your local logs to a private Grafana instance or Grafana Cloud.

Promtail allows you to write powerful and complex pipelines that can transform your logs prior to export to your Loki instances. For example, perhaps you want to use a regex to extract your log lines into searchable fields, or add a new label to your logs, or drop log lines that you don't need.

And while you may think that you need to deploy updates to your Promtail configuration in order to test them, you can actually test these locally first with a few simple steps!

In this example, we'll drop log lines that we don't need in Loki.

Install Promtail #

If you're on macOS, you can install with brew.

brew install promtail

There are also installation options for other operating systems.

Create Example Logs #

Create a file that contains example logs lines that you want to test. We'll save this as logs.txt.


Create a Promtail Configuration File #

This is a regular Promtail configuration file with the caveat that you can only have one scrape config. Also, if you need to add labels to log lines, you can with a static_configs section.

- url: http://localhost:3100/loki/api/v1/push
- job_name: testing-my-job-drop
- match:
selector: '{job="my-job"}'
- json:
- drop:
source: "request_path"
expression: "(/healthcheck|/metrics)"
drop_counter_reason: my_job_health_or_metrics
- labels:
job: my-job

Here we're saying only match on logs that have the my-job job label. We add that label to the log lines with the static_configs section. We then extract request_path from the JSON log line and then drop all lines that have /healthcheck or /metrics. Save this one as promtail.yml.

Run Promtail in Dry Run Mode #

Now that we're setup, we can run Promtail locally and test our config!

$ cat logs.txt | promtail --config.file ./promtail.yml --stdin --dry-run --inspect
Clients configured:
url: http://localhost:3100/loki/api/v1/push
batchwait: 1s
batchsize: 1048576
follow_redirects: false
min_period: 500ms
max_period: 5m0s
max_retries: 10
timeout: 10s
tenant_id: ""
stream_lag_labels: filename

level=info ts=2022-03-19T15:24:24.318359Z caller=server.go:260 http=[::]:80 grpc=[::]:9095 msg="server listening on addresses"
[inspect: json stage]:
+: /actual-endpoint
2022-03-19T09:24:24.317955-0600{job="my-job"} {"remote_ip":"","request_id":"-","response_code":"200","request_method":"POST","request_path":"/actual-endpoint","request_querystring":"","request_timetaken":"6120","response_length":"22"}
level=info ts=2022-03-19T15:24:24.318525Z caller=main.go:119 msg="Starting Promtail" version="(version=, branch=, revision=)"
[inspect: json stage]:
+: /metrics
[inspect: json stage]:
+: /healthcheck

There it is!

We can see that Promtail extracted the request_path from our JSON log lines and then drops the lines that contain /healthcheck or /metrics leaving us with the single /actual-endpoint line.

Conclusion #

So, with a little setup and a carefully crafted Promtail configuration, you can test your pipeline stages locally and feel confident that your production setup will work as expected.