How to send metrics and logs to your Cockpit

This page shows you how to send data such as metrics and logs to your Cockpit. You can push metrics with any Prometheus Remote Write compatible agent such as the Prometheus, Grafana or OpenTelemetry Collector agents. You can push logs with any Loki compatible agent such as the Promtail, Fluentd, Fluent Bit or Logstash agents.

Before you startLink to this anchor

To complete the actions presented below, you must have:

• A Scaleway account logged into the console
• Owner status or IAM permissions allowing you to perform actions in the intended Organization
• Configured an agent. Read our dedicated documentation to find out how to configure the Grafana Alloy agent
• Created an API key and retrieved your API secret key

Send metrics and logs to your CockpitLink to this anchor

1. Click Cockpit in the Monitoring section of the console side menu. The Cockpit overview page displays.

2. Create a token with the Push permission for metrics and logs.

   Important

   • Make sure that you create your token in the same region as the data sources you want to use it for.
   • Make sure that you copy your token’s secret key and save it before you exit the page.

3. Configure your applications to send metrics and logs to your Cockpit endpoints, using your token’s secret key for authentication.

   Important

   Retrieve your metrics’ and logs’ push endpoints in the Data sources tab of the Scaleway console.

4. Log in to your Grafana account.

5. Create your Grafana dashboard and start monitoring your data.

Sending Python logs to CockpitLink to this anchor

The section below provides you with a Python script that sends test log messages to Cockpit. It is a safe way to ensure that your custom log setup works before integrating real application logs.

1. Open a terminal and run pip install python-logging-loki or pip3 install python-logging-loki if you are using Python 3, to install the logging_loki package to send your Python logs.
2. Create a Cockpit token with the Push permission for logs.
3. Create a python_logs.py Python file and paste the following code snippet into it. Make sure that you replace 000avb0d-34ae-66hh-643b-f9e0n3k17773 with the ID of your logs data source, $SCW_API_SECRET_KEY with your API secret key and $COCKPIT_TOKEN_SECRET_KEY with your Cockpit token’s secret key.

import logging
import logging_loki
import time
handler = logging_loki.LokiHandler(
    url="https://000avb0d-34ae-66hh-643b-f9e0n3k17773.cockpit.fr-par.scw.cloud/loki/api/v1/push",
    tags={"job": "logs_from_python"},
    auth=("$SCW_API_SECRET_KEY", "$COCKPIT_TOKEN_SECRET_KEY"),
    version="1",
)
logger = logging.getLogger("my-first-python-logger")
logger.addHandler(handler)
def main():
    # Start generating log messages
    for i in range(10):
        logger.error(f"Error message {i}")
        time.sleep(1)
if __name__ == "__main__":
    main()

This script sets up a logger that sends error log messages to the Loki endpoint in Cockpit. It then generates 10 error log messages with a one-second delay between each message. These log messages are sent to Cockpit and can be viewed and analyzed via the Grafana dashboard we will create in the next part.

Creating a Grafana dashboard and visualizing your Python logsLink to this anchor

1. Log into Grafana using your Grafana credentials.

2. Click Toggle menu, then click Dashboards.

3. Click New in the top right corner of your screen, then Import.

4. Upload the JSON file of the Grafana Log Viewer dashboard for Loki or enter 13639 in the Import via grafana.com field.

5. Click Load.

6. Enter a name for the dashboard or leave the default name.

7. Select a folder in which to store the dashboard.

8. Select your custom logs data source in the Loki drop-down. Make sure that the data source you select is in the fr-par region.

9. Click Import.

10. Open a terminal and enter the following command.

    python3 python_logs.py
    

11. Click the Home icon > Dashboards. Your Logs / App dashboard should display.

12. Click the Logs / App dashboard. You should see 10 log lines labeled Error message 0 through Error message 9, one per second. This confirms that your setup is working.

    If your logs do not display, try changing the time range to match the time at which you ran your Python file.

    

13. Click the View in Explore button for more configuration options.