How can metrics be collected from client applications for the Prometheus system?

To configure metrics collection from client applications in the Prometheus system, exporters or client libraries are usually used, which integrate into the application and provide metrics in a format understandable by Prometheus.

Main steps:

1. Connect the Prometheus client library in the application (for example, prometheus_client for Python).
2. Define and update the necessary metrics (counters, histograms, timers).
3. Run an HTTP server that will serve metrics at a specific endpoint (e.g., /metrics).
4. In the Prometheus configuration, add a job specifying the application's address and port for metrics collection.

Example in Python:

from prometheus_client import start_http_server, Counter
import time

REQUESTS = Counter('app_requests_total', 'Total number of requests')

if __name__ == '__main__':
    start_http_server(8000)  # Metrics are available at http://localhost:8000/metrics
    while True:
        REQUESTS.inc()  # Increment the counter
        time.sleep(1)

Thus, Prometheus will regularly poll this endpoint and collect metrics.