Prometheus & Grafana Bootcamp: Monitoring for DevOps & SRE

Prometheus provides powerful monitoring and alerting by scraping metrics from operating systems and applications via exporters and client libraries, pulling data over HTTP to the Prometheus server.

Learn how Prometheus scrapes metrics via http endpoints, stores data in a time series database, and uses counters and gauges, with histogram and summary metrics discussed later.

Explore Prometheus metric types, including counter, gauge, summary, and histogram, and learn to calculate quantiles, distributions, and latency with bucketed data.

Explore Prometheus installation options—from docker containers and precompiled binaries to Kubernetes helm charts—installing on a Unix cloud server to mirror production, with a basic Prometheus UI.

Create a cloud machine on DigitalOcean with a $100 credit and install Prometheus on Ubuntu 20.04, using SSH key or password authentication and the browser console.

Install Prometheus on a unix/linux machine by downloading the official binary from prometheus.io, extracting it, and setting up the server on a DigitalOcean instance.

Launch the Prometheus server and access its user interface on port 9090 to explore graphs, alerts, status, and targets, while learning core metrics like up.

Explore Prometheus default configuration, including global, scrape config, and alerting sections, and how the YAML defines 15s scrape and evaluation intervals, 10s timeouts, rules, and local targets.

launch Prometheus as a background process using nohup, redirect logs to a file, and monitor it with ps, then access Prometheus in a browser while the terminal stays free.

Learn how Prometheus exporters extend existing systems by translating metrics into Prometheus format, install and configure exporters (like node exporter), and set up jobs to scrape data.

Configure and monitor Linux servers with node exporter and Prometheus, exposing system metrics on port 910, scraping via a Prometheus job, and visualizing in Prometheus UI.

Install and configure the node exporter on a linux machine and run it on port 9100. Update prometheus.yaml to add node exporter job for localhost:9100 and verify metrics in Prometheus.

Learn how the WMI exporter—an unofficial Prometheus exporter for Windows—enables monitoring Windows via Windows Management Instrumentation metrics, installed via the amd64 exe, for local or network scraping.

Explore monitoring MySQL databases with Prometheus using the MySQL server exporter, configuring a scrape job, and comparing database metrics to system metrics under varying loads.

Download and extract the MySQL exporter on the target machine, create the exporter user with replication and select privileges, and configure credentials via environment variables or a cnf file.

Reload the Prometheus configuration without restarting by enabling the web.enable.lifecycle flag and posting to /-/reload via curl on port 9090, applying changes to exporters and targets live.

Demonstrates installing the MySQL exporter, including downloading, extracting, and moving the binary to /usr/local/bin. Creates an exporter user, grants privileges, and sets up Prometheus integration.

Configure the MySQL exporter as a systemd service by creating a cnf and service file, then enable, start, and verify metrics on port 9104 for Prometheus.

Discover how the black box exporter enables Prometheus to monitor network endpoints by probing HTTP, HTTPS, DNS, and ICMP with configurable modules, improving uptime and visibility.

Download and configure the black box exporter for Prometheus, extract the tar, and run it on port 9015 to expose metrics.

Explore the http probe module in black box exporter, and see how Prometheus targets and modules trigger probes, view results, and debug with debug=true to resolve IPv6 related failures.

Learn to configure an http probe for IPv4 using a custom module in the black box exporter, reload Prometheus, and observe successful probes, DNS lookups, status 200s, and redirects.

Demonstrate using the http probe module to validate page content with a regex, configure targets in black box exporter, and monitor ssl expiry via epoch timestamps for proactive alerts.

Explore the tcp probe in Prometheus' black box exporter, which validates a tcp connection and closes after success, without ssl certificate verification, contrasting with the http probe.

Learn how to configure a DNS probe module in the Prometheus black box exporter to verify DNS servers by querying google.com, switching from UDP to TCP, and examining DNS responses.

Learn to configure Prometheus to scrape a black box exporter, define targets, adjust the metric path to /prop, and apply relabeling for correct target scraping.

Learn how relabeling in Prometheus enables scraping data from the black box exporter for Prometheus IO targets, by mapping source labels to target labels and configuring modules.

Learn how to install Prometheus on Kubernetes using the Prometheus operator and helm chart, enable data scraping via service discovery, and monitor Kubernetes components with node exporter, alertmanager, and Pushgateway.

Install Prometheus in Kubernetes using the Prometheus community helm chart, add the repo, install with helm, and verify resources like Prometheus server, alert manager, kube state metrics, and node exporter.

Learn methods to access Prometheus on a Kubernetes cluster, including port forwarding, node port and load balancer services, and viewing Prometheus targets, pods, and metrics.

Explore Prometheus on Kubernetes and learn which native Kubernetes metrics are available, how to view pod, service, and API server metrics, and prepare for Grafana visualization.

Learn how to use Prometheus client libraries to instrument your application, understand official vs unofficial libraries, and explore core metric types: counter, gauge, histogram, and summary.

Set up a basic Python web server and instrument it with the Prometheus client library, using Python 3 and Visual Studio Code.

Expose metrics from a Python app using the Prometheus client library, install via pip, start an HTTP server on port 5001, and observe metrics like python_gc_object_total and python_info.

Monitor a custom Python app with Prometheus by deploying the app on the Prometheus server, exposing metrics, and configuring the Prometheus YAML target for scraping.

Expose a counter metric in a custom application and monitor it with Prometheus. Define and increment the metric, and view app request counts and creation timestamps on Prometheus graphs.

Add labels to exposed Prometheus metrics to capture per-endpoint request data by extending counter metrics with label names, attaching labels in code, and visualizing them in Prometheus.

Learn how to expose gauge metrics with Prometheus in a Python app, tracking live in-progress requests and last served time, with inc, dec, set, and track_in_progress annotations.

Expose the summary metric in Prometheus to measure request latency, implementing a summary metric in Python, calculating latency with start and end times, and using observe to collect data.

Learn how to expose and compare histogram metrics in Prometheus, including bucket interpretation, latency measurement, and integrating histogram alongside summary metrics using Python decorators.

Set up a Linux machine on DigitalOcean, install Docker, and run a dockerized Python Flask web app. Learn how to monitor the application with Prometheus in a Docker environment.

Instrument a dockerized Python Flask app with Prometheus metrics by creating an app.py, exposing the /metrics endpoint, and building a docker image using a tailored dockerfile.

Create and run a docker image for a Python web app; build from a dockerfile, fix the filename case, run on port 5001, and access metrics at /metrics for Prometheus.

Monitor a dockerized application with Prometheus by updating the Prometheus YAML to a custom docker app and restarting Prometheus, then validate metrics like app_request_counter_total from the target IP.

Set up and instrument a Go application by creating a sample http server, initializing go mod init, adding dependencies with go get, and running on port 8000.

Instrument a go app with a Prometheus counter, expose /metrics, and configure Prometheus to scrape the app, enabling real-time visibility of request counts.

Expose a gauge metric in a Go app by replacing the counter, incrementing at request start and decrementing at end, with a delay to observe in progress requests.

Expose a summary metric in a Go app by creating a summary vector, updating its name and help, and adding route middleware to measure latency per path for Prometheus.

Instrument a Go app with a histogram metric in Prometheus, replacing summary, to measure per-path latency and visualize execution times in the metrics page.