Monitoring Prometheus metric with AppDynamics using OpenTelemetry

[Abhi.Bajaj]

In this Knowledge Base Article, we’ll walk you through the process of collecting Prometheus metrics from a Python application, forwarding them to AppDynamics using OpenTelemetry, and visualizing them for effective monitoring.

Setting up the Python Application

Let’s start with creating a Python application that generates Prometheus metrics. We’ll use the prometheus_client library to create and expose these metrics. If you haven't installed the library, you can do so with:

pip3 install prometheus_client

Now, let’s dive into the Python script:

import random
import time
from prometheus_client import start_http_server, Counter, Summary

# Define Prometheus metrics
some_counter = Counter(name="myapp_some_counter_total", documentation="Sample counter")
request_latency = Summary(name="myapp_request_latency_seconds", documentation="Request latency in seconds")

def main() -> None:
    start_http_server(port=9090)
    while True:
        try:
            # Simulate application logic here
            process_request()
            time.sleep(5)  # Sleep for a few seconds between metric updates
        except KeyboardInterrupt:
            break

def process_request():
    # Simulate processing a request and record metrics
    with request_latency.time():
        random_sleep_time = random.uniform(0.1, 0.5)
        time.sleep(random_sleep_time)
    some_counter.inc()

if __name__ == "__main__":
    main()

This Python script sets up a simple HTTP server on port 9090 and generates two Prometheus metrics: myapp_some_counter_total and myapp_request_latency_seconds.

To produce load:

curl -v http://localhost:9090

The logs will look like:

*   Trying 127.0.0.1:8081...
* Connected to localhost (127.0.0.1) port 8081 (#0)
> GET / HTTP/1.1
> Host: localhost:8081
> User-Agent: curl/7.81.0
> Accept: */*
>
* Mark bundle as not supporting multiuse
* HTTP 1.0, assume close after body
< HTTP/1.0 200 OK
< Date: Thu, 07 Dec 2023 15:58:00 GMT
< Server: WSGIServer/0.2 CPython/3.10.12
< Content-Type: text/plain; version=0.0.4; charset=utf-8
< Content-Length: 2527
<
# HELP python_gc_objects_collected_total Objects collected during gc
# TYPE python_gc_objects_collected_total counter
python_gc_objects_collected_total{generation="0"} 371.0
python_gc_objects_collected_total{generation="1"} 33.0
python_gc_objects_collected_total{generation="2"} 0.0
# HELP python_gc_objects_uncollectable_total Uncollectable objects found during GC
# TYPE python_gc_objects_uncollectable_total counter
python_gc_objects_uncollectable_total{generation="0"} 0.0
python_gc_objects_uncollectable_total{generation="1"} 0.0
python_gc_objects_uncollectable_total{generation="2"} 0.0
# HELP python_gc_collections_total Number of times this generation was collected
# TYPE python_gc_collections_total counter
python_gc_collections_total{generation="0"} 40.0
python_gc_collections_total{generation="1"} 3.0
python_gc_collections_total{generation="2"} 0.0

Deploying OpenTelemetry Collector

To collect and forward metrics to AppDynamics, we’ll use OpenTelemetry Collector. This component plays a vital role in gathering metrics from various sources and exporting them to different backends. In this case, we’ll configure it to forward metrics to AppDynamics.

Installing OpenTelemetry Collector on Ubuntu

1. Make sure you’re on an Ubuntu machine. If not, adjust the installation instructions accordingly.
2. Install OpenTelemetry Collector:
   https://medium.com/@abhimanyubajaj98/linux-host-monitoring-with-appdynamics-deploying-opentelemetry-...
3. For this tutorial, we will edit /opt/appdynamics/appdynamics.conf and add another variable called:

APPD_OTELCOL_EXTRA_CONFIG=--config=file:/opt/appdynamics/config.yaml

Our appdynamics.conf file will look like:

APPD_OTELCOL_CLIENT_ID=<client-id>
APPD_OTELCOL_CLIENT_SECRET=<client-secret>
APPD_OTELCOL_TOKEN_URL=<tenant-url>
APPD_OTELCOL_ENDPOINT_URL=<tenant_endpoint>
APPD_LOGCOL_COLLECTORS_LOGGING_ENABLED=true
APPD_OTELCOL_EXTRA_CONFIG=--config=file:/opt/appdynamics/config.yaml

We will create the config.yaml like below.

Configuring OpenTelemetry Collector

Create an config.yaml configuration file with the following content:

extensions:
  oauth2client:
    client_id: "${env:APPD_OTELCOL_CLIENT_ID}"
    client_secret: "${env:APPD_OTELCOL_CLIENT_SECRET}"
    token_url: "${env:APPD_OTELCOL_TOKEN_URL}"
  health_check:
    endpoint: 0.0.0.0:13133
  zpages:
    endpoint: 0.0.0.0:55679

receivers:
  otlp:
    protocols:
      grpc:
        endpoint: 0.0.0.0:4317
      http:
        endpoint: 0.0.0.0:4318
  prometheus:
    config:
      scrape_configs:
        - job_name: 'prometheus'
          static_configs:
            - targets: ['localhost:9090']

processors:
  # defaults based on perf testing for k8s nodes
  batch:
    send_batch_size: 1000
    timeout: 10s
    send_batch_max_size: 1000

  memory_limiter:
    check_interval: 5s
    limit_mib: 1536

exporters:
  otlphttp:
    retry_on_failure:
      max_elapsed_time: 180
    metrics_endpoint: "${env:APPD_OTELCOL_ENDPOINT_URL}/v1/metrics"
    traces_endpoint: "${env:APPD_OTELCOL_ENDPOINT_URL}/v1/trace"
    logs_endpoint: "${env:APPD_OTELCOL_ENDPOINT_URL}/v1/logs"
    auth:
      authenticator: oauth2client

service:
  telemetry:
        logs:
          level: debug
  extensions: [zpages, health_check, oauth2client]
  pipelines:
    metrics:
      receivers: [prometheus, otlp]
      processors: [memory_limiter, batch]
      exporters: [otlphttp]
    traces:
      receivers: [otlp]
      processors: [memory_limiter, batch]
      exporters: [otlphttp]

In this configuration file, we have set up the OpenTelemetry Collector to receive metrics from the Prometheus receiver and export them to AppDynamics using the OTLP exporter.

Forwarding Metrics to AppDynamics

With the OpenTelemetry Collector configured, it will now collect metrics from your Python application and forward them to AppDynamics. This seamless integration enables you to monitor your application’s performance in real-time.

Monitoring Metrics in AppDynamics

You can use UQL to query the metrics.
This is a very basic example, you can create attributes.

To learn more about the AppDynamics metric model, check out this AppDynamics Docs page