
Already know some of this? A 2-minute test skips the lessons you know.
Take the placement testA free 32-lesson course built from Boot dev's 7h 14m of YouTube video. 5 days at 90 minutes a day, with study notes and a quiz in every lesson. A 2-minute placement test skips what you already know.
Taught by Boot dev on YouTube. Videos play in YouTube's own player, so the creator keeps every view.
Understand the course goals, set up the Boot.dev CLI tool, and authenticate to track progress.
Clone the Linko URL shortener starter repository, run it locally, and verify it serves requests.
Define observability and distinguish its three pillars—logs, metrics, and traces—plus alerting.
Replace all fmt.Printf/Println calls with log package equivalents and remove manual newlines.
Instantiate custom log.Logger objects, set output destinations (file, stderr), prefixes, and flags.
Pass loggers through the application via dependency injection instead of using global variables.
Switch to slog for structured logging, use log levels (debug, info, warn, error), and filter by level.
Emit structured logs with slog.Info/Error, add key-value attributes, and output JSON to files.
Configure slog.MultiHandler to send debug+ logs to stderr and info+ logs to a file simultaneously.
Add timestamps, build version/time via -ldflags, and propagate request-scoped context through middleware.
Wrap errors with fmt.Errorf(%w), add stack traces via pkg/errors, and log errors only at the decision point.
Separate error message and stack trace into structured slog.Group fields for better queryability.
Detect errors.MultiError, unwrap it, and log each contained error under numbered keys in a top-level group.
Implement a replaceAttr filter to redact sensitive keys (password, api_key) and obfuscate IPs/URLs with credentials.
Add post-write log filtering to catch accidentally logged secrets, but treat it as a last line of defense.
Replace manual file/buffered writer with Lumberjack for automatic size- and time-based log rotation.
Distinguish metric types and choose the right one (rate, errors, duration) for a given observability question.
Launch the observability stack with Docker Compose, add Grafana, and import a host system dashboard.
Instrument Linko with promhttp.Handler, configure Prometheus to scrape it, and verify Go runtime metrics appear.
Create Prometheus CounterVecs for request rate, errors, and duration with labels; visualize in Grafana.
Create a Grafana alert rule, tune threshold/window/pending period, and shift from 401 spikes to 500 error spikes.
Generate CPU and heap profiles under load, visualize with Graphviz, and identify hotspots (blowfish encrypt, strings.Builder).
Snapshot goroutines with pprof, interpret runtime.park/select/handler states, and recognize leak patterns.
Initialize a TracerProvider, add otelhttp middleware for root spans, and create child spans in handlers.
Query Jaeger for slow traces, inspect span timelines, and pinpoint the exact function causing latency.