Performance Engineering of Software Systems Course

Lesson 1 • Performance Engineering Workflow

Presents the iterative measure-analyse-improve cycle as the core workflow. Orients students to the structure of the entire course.

Lesson 2 • Performance Goals and SLOs

Teaches how to translate business requirements into measurable performance targets. Establishes the goal-setting framework applied in every later chapter.

Lesson 3 • System Resource Model

Maps CPU, memory, disk, and network to performance constraints. Provides a shared model used throughout all subsequent chapters.

Lesson 4 • Key Performance Metrics and Indicators

Introduces latency, throughput, error rate, and saturation as primary signals. Connects metrics to user experience and business outcomes.

Lesson 1 • Profiling CPU and Memory

Introduces sampling and tracing profilers for CPU and heap analysis. Builds the skill of identifying hot paths and memory leaks from profiles.

Lesson 2 • Instrumentation Techniques

Covers manual, automatic, and bytecode instrumentation approaches. Connects instrumentation choices to overhead and data fidelity trade-offs.

Lesson 3 • Dashboards and Alerting Design

Guides creation of actionable dashboards and low-noise alert policies. Ensures collected data drives decisions rather than generating alert fatigue.

Lesson 4 • Observability Pillars: Metrics, Logs, Traces

Distinguishes the three telemetry types and their complementary roles. Anchors the chapter by showing when each pillar is most valuable.

Lesson 1 • Load Test Types and Objectives

Distinguishes baseline, stress, soak, spike, and breakpoint tests. Aligns each test type to specific performance questions.

Lesson 2 • Workload Characterisation

Analyses production traffic patterns to build representative workload models. Connects real usage data to synthetic test scenarios.

Lesson 3 • Load Generator Design

Covers scripting realistic virtual users and avoiding coordinated omission. Ensures the load generator itself does not become the bottleneck.

Lesson 4 • Analysing Load Test Results

Teaches systematic result analysis, including percentile comparison and bottleneck identification. Bridges raw test output to actionable findings.

Lesson 1 • CPU Bottleneck Diagnosis

Identifies CPU-bound conditions through profiling, run-queue depth, and scheduling analysis. Connects CPU symptoms to root causes in application code.

Lesson 2 • Concurrency and Lock Contention

Diagnoses thread contention, deadlocks, and lock convoy effects. Applies Amdahl's Law to quantify parallelism limits.

Lesson 3 • Memory and GC Bottleneck Diagnosis

Diagnoses heap pressure, garbage collection pauses, and memory leaks. Links memory behaviour to latency spikes and throughput degradation.

Lesson 4 • I/O and Storage Bottleneck Diagnosis

Covers disk I/O wait, buffer cache behaviour, and storage queue depth analysis. Connects I/O patterns to application-level read/write strategies.

Lesson 1 • Algorithmic Complexity and Data Structures

Reviews Big-O analysis and selects optimal data structures for performance-critical paths. Establishes the highest-leverage optimisation layer.

Lesson 2 • Caching Strategies

Covers in-process, distributed, and CDN caching with eviction policies and invalidation. Directly reduces latency and backend load.

Lesson 3 • Serialisation and Protocol Optimisation

Compares serialisation formats and protocol choices for throughput and latency. Reduces wire overhead and CPU cost of data encoding.

Lesson 4 • Database Query Optimisation

Covers query plan analysis, indexing strategies, and N+1 query elimination. Addresses the most common application-level performance bottleneck.

Lesson 1 • Resilience and Graceful Degradation

Designs circuit breakers, bulkheads, and load shedding to maintain performance under failure. Prevents cascading failures from collapsing throughput.

Lesson 2 • Capacity Planning and Modelling

Teaches demand forecasting, resource modelling, and headroom calculation. Translates growth projections into infrastructure decisions.

Lesson 3 • Scalability Models and Laws

Applies Amdahl's and Gunther's Universal Scalability Law to predict system limits. Provides quantitative tools for architecture decisions.

Lesson 4 • Horizontal Scaling Patterns

Covers stateless service design, load balancing algorithms, and session affinity trade-offs. Enables linear capacity growth through replication.

Lesson 1 • Content Delivery and Edge Performance

Uses CDN caching, edge compute, and DNS optimisation to reduce global latency. Extends performance engineering to the client-facing delivery layer.

Lesson 2 • Cloud Resource Performance Tuning

Optimises instance types, storage tiers, and managed service configurations for cost-performance. Addresses cloud-specific throttling and quota limits.

Lesson 3 • Network Latency in Distributed Systems

Quantifies the impact of network hops, serialisation, and fan-out on end-to-end latency. Motivates co-location, batching, and protocol choices.

Lesson 4 • Microservices Performance Patterns

Covers service decomposition trade-offs, API gateway overhead, and inter-service call optimisation. Applies distributed tracing to cross-service diagnosis.

Lesson 1 • Production Performance Monitoring

Establishes real-user monitoring, synthetic probes, and anomaly detection in production. Closes the feedback loop between deployment and measurement.

Lesson 2 • Incident Response for Performance Issues

Applies a structured runbook approach to diagnosing and resolving live performance incidents. Integrates observability skills into on-call practice.

Lesson 3 • Performance Regression Detection

Designs automated gates that catch regressions before production deployment. Connects measurement from earlier chapters to pipeline enforcement.

Lesson 4 • Performance Budgets and Governance

Sets team-level performance budgets and review processes to sustain gains over time. Institutionalises performance as a first-class engineering concern.