Prometheus

If this were a book, this page is the table of contents. Each Part below is a chapter; each chapter links out to the concepts, designs, and platform notes that already exist elsewhere in this wiki instead of duplicating them. This book is intentionally deeper than the PCA syllabus — several Parts (production internals, PCA prep, MAANG interview banks) are scaffolded ahead of their source material and marked (stub) below; they’ll fill in as that material gets written.

Parts

00 — Monitoring Foundations

The vocabulary this whole book assumes: what a time series actually is, and where Prometheus sits relative to the rest of the observability landscape.

• 1 — Why Monitoring Exists — (stub) — see What Observability Actually Means and SLOs & Error Budgets for the 3-pillars/SLI-SLO-SLA material this chapter would otherwise duplicate
• 2 — Time Series Fundamentals — samples, labels, dimensions, and metadata as Prometheus’s on-the-wire data model
• 3 — Prometheus in the Observability Ecosystem — architecture flow, CNCF history, and where Mimir/Cortex/Thanos/VictoriaMetrics fit

01 — Prometheus Architecture

The server itself: how it’s installed and run, why it pulls instead of accepting pushes, and how a metric moves from a target to a dashboard.

• 1 — Prometheus Components — Server/TSDB/Scrape Manager/Rule Engine, plus real bare-metal, systemd, and Docker install steps
• 2 — Pull Model Deep Dive — why pull, the Pushgateway escape hatch for batch jobs; see Push vs. Pull Ingestion for the general architecture treatment
• 3 — Data Flow — instrumentation through dashboards, tying every later Part together

02 — Prometheus Data Model

Counters, gauges, histograms, summaries, labels, and the on-disk storage engine underneath them.

• 1 — Metrics Deep Dive — the four metric types via a real hands-on lab (this is the canonical home for this book’s demo environment — other chapters link here rather than re-embedding it)
• 2 — Labels and Cardinality — label mechanics, relabeling, and the storage/performance cost of cardinality; see Cardinality and Label & Attribute Schema Design for budget/governance depth
• 3 — TSDB Internals — (stub) — see Metrics Storage (TSDB) for the WAL/compaction internals this chapter doesn’t cover yet

03 — Instrumentation

Getting metrics out of a system in the first place — client libraries, exporters, and writing your own instrumentation.

• 1 — Client Libraries — (stub)
• 2 — Exporters — Node Exporter and Windows Exporter install, plus Docker/cAdvisor container monitoring
• 3 — Custom Instrumentation — (stub)

04 — Service Discovery

Finding what to scrape without hand-maintaining a static target list.

• 1 — Discovery Mechanisms — static configs, file-based SD, and DNS SD
• 2 — Kubernetes Discovery — (stub)
• 3 — Cloud Discovery — (stub)

05 — PromQL Masterclass

The query language in full — this Part is the deepest and most complete in the book.

• 1 — PromQL Fundamentals — data types, selectors, matchers, and running queries outside the Prometheus UI
• 2 — PromQL Functions — the rate family, math/date-time/sorting functions, and histogram_quantile() mechanics
• 3 — Aggregation Operators — sum/avg/count/topk/bottomk/quantile/stddev/stdvar, by/without
• 4 — Vector Matching — on/ignoring, one-to-one vs. many-to-one, and operator precedence
• 5 — Advanced PromQL — offset/@ modifiers, subqueries, and this book’s canonical recording-rule syntax reference

06 — Alerting

Turning PromQL expressions into pages — recording rules, alert rules, and Alertmanager.

• 1 — Recording Rules — why and when, cross-linking Advanced PromQL for the syntax itself
• 2 — Alerting Rules — the pending → firing lifecycle and a real templated alert
• 3 — Alertmanager — group_wait/group_interval/ repeat_interval grouping and dedup mechanics; see Alerting & Alert Routing for routing/escalation depth

07 — Production Prometheus

Running Prometheus past the point where one server on one disk is enough.

• 1 — Scaling Prometheus — (stub)
• 2 — Long-Term Storage — why Thanos/Cortex/Mimir/VictoriaMetrics exist; see Mimir and Scaling Each Layer
• 3 — Performance Tuning — (stub)
• 4 — High Availability — (stub)

08 — Operating Prometheus

Day-2 operations: Kubernetes deployment patterns, hardening, and troubleshooting.

• 1 — Kubernetes Best Practices — (stub)
• 2 — Security — a full TLS + basic-auth walkthrough (self-signed certs, htpasswd, tls_server_config)
• 3 — Troubleshooting — (stub)

09 — Prometheus Certified Associate (PCA)

Certification-specific prep, built on top of the material already covered above.

• 1 — PCA Exam Objectives — (stub)
• 2 — Hands-on Labs — an ordered lab path through this book’s existing chapters
• 3 — Practice Exams — (stub)

10 — MAANG Interview Preparation

System-design framing and the “why,” not just the “how.”

• 1 — Prometheus System Design — (stub)
• 2 — Interview Questions — (stub)
• 3 — Deep Dive Discussions — why pull instead of push, why not SQL
• 4 — Real Production Architectures — (stub)

11 — Appendices

Quick-reference material distilled from the chapters above.

• 1 — PromQL Cheat Sheet — grouped CPU/memory/disk/network/swap/inode/TCP query reference
• 2 — Recording Rule Cookbook — (stub — this book’s only recording-rule worked examples already live in Advanced PromQL)
• 3 — Alert Rule Cookbook — (stub — only one worked alert example exists in this book so far)
• 4 — Exporter Catalog — Node/Windows Exporter metric lookup table
• 5 — Prometheus Configuration Reference — field-by-field prometheus.yml reference
• 6 — Common Anti-Patterns — three traced-to-source mistakes: high-cardinality labels, metric-naming violations, misaligned histogram buckets
• 7 — PCA Exam Cheat Sheet — (stub)
• 8 — Interview Cheat Sheet — (stub)

Metadata

Author	Amit Singh
Scope	prometheus