Java Developer Interview Prep Guide

Virtual threads are lightweight, JVM-managed threads that are cheap to create (millions possible) and block cheaply. Platform threads map 1:1 to OS threads and are expensive. Use virtual threads for I/O-bound workloads like HTTP clients, database calls, and file operations. Keep platform threads for CPU-bound work. Discuss how virtual threads simplify concurrent code by replacing reactive programming with straightforward blocking style.

Use Kafka for event streaming with idempotent producers and exactly-once semantics via transactions. Spring Boot with virtual threads for handling concurrent requests. Discuss the outbox pattern for reliable event publishing from a database, consumer group scaling, dead letter topics for failed messages, and how to handle duplicate detection with idempotency keys stored in Redis.

G1 is the default: good balance of throughput and latency with region-based collection. ZGC offers sub-millisecond pause times regardless of heap size, ideal for latency-sensitive applications. Shenandoah is similar to ZGC but with different algorithms. Choose G1 for general workloads, ZGC for low-latency requirements (trading services, real-time APIs), and discuss how to monitor GC performance with GC logs, JFR, and tools like GCViewer.

Check GC logs first for stop-the-world pauses. Use JFR (Java Flight Recorder) for low-overhead production profiling. Look at thread dumps for lock contention or thread pool exhaustion. Check connection pool metrics (HikariCP). Investigate OS-level issues like swap usage or CPU throttling in containerized environments. Discuss how you would set up monitoring to catch this proactively.

Use a circular array with head and tail pointers, synchronized methods or ReentrantLock with two Conditions (notFull, notEmpty). Show proper wait/notify pattern with while loops (not if) to handle spurious wakeups. Discuss why you would use ReentrantLock with Condition over synchronized in production code: fairness option, tryLock with timeout, and multiple conditions per lock.

Auto-configuration uses @Conditional annotations (ConditionalOnClass, ConditionalOnMissingBean) and spring.factories or AutoConfiguration.imports to register configuration classes. A custom starter contains: auto-configuration module with @Configuration classes, a starter module with dependencies, and @ConfigurationProperties for externalized configuration. Discuss how Spring Boot orders auto-configurations and resolves conflicts.

Describe migrating from Java 8 to 21, monolith to microservices, or XML configuration to Spring Boot. Discuss the strangler fig pattern for incremental migration, how you maintained backward compatibility, testing strategy during migration, and specific challenges like library incompatibilities or behavioral changes. Include metrics: deployment frequency improvement, latency reduction, or developer productivity gains.

Unit tests with JUnit 5 and Mockito for business logic. Integration tests with @SpringBootTest and Testcontainers for real database and Kafka testing. Contract tests with Spring Cloud Contract or Pact for API compatibility. Discuss test slicing (@WebMvcTest, @DataJpaTest) for faster integration tests, test fixtures, and how to maintain test data across environments.