*We know each team has their own needs and specifications. That is why we can modify the training outline per need.
Module 1: Performance mindset and honest measurement
- Throughput vs latency vs tail latency and why they differ
- Benchmark pitfalls warmup, JIT, constant folding, and fair comparisons
- Micro vs macro benchmarks with JMH at a glance
- Choosing target SLOs and a repeatable test plan
Module 2: Profiling CPU and allocations
- Flame graphs and stack sampling with async profiler or JFR
- Hot methods, inlining, boxing, and allocation hot spots
- Object lifetime basics escape analysis and scalar replacement ideas
- Reading JIT compilation logs at a beginner level
Module 3: Memory and garbage collection tuning
- Heap layout young, old, metaspace and what to watch
- Collector choices G1, ZGC, Shenandoah and selection tips
- Key GC knobs pause goals, regions, initiating heap occupancy
- Reading GC logs and linking pauses to allocation rate
Module 4: Concurrency and synchronization costs
- Contention sources locks, atomics, false sharing, and queues
- Thread pools, virtual threads overview, and backpressure signals
- Timeouts, retries, and circuit breakers to protect tail latency
- Safe patterns for caching and avoiding stampedes
Module 5: I or O, databases, and downstream effects
- HTTP clients, connection pools, timeouts, and DNS caching
- JDBC and R2DBC basics fetch size, batching, and N plus 1 traps
- Serialization costs JSON, binary options, and payload shaping
- Async pipelines and when to keep it simple
Module 6: JVM and container settings
- Memory sizing heap vs native and container awareness
- CPU limits, schedulers, and impact of throttling
- Class data sharing, tiered compilation, and startup tradeoffs
- Safe defaults for production flags and a small checklist
Module 7: Observability for speed
- Metrics that matter request, pool, GC, and queue signals
- Tracing to spot slow hops and head of line blocking
- Log hygiene for timing without noise
- Building a simple performance dashboard
Module 8: Tuning workflow and playbooks
- Triage steps reproduce, measure, change, verify
- Rollout strategies canary and feature flags for risky changes
- Common tuning recipes allocation cuts, pool tuning, GC guardrails
- Ninety day plan and a pocket checklist
