D
DevPrepSystematic Prep
JAVA · HARD CODEX

Java Interview Questions — Hard

Deep runtime internals, memory models, distributed design, concurrency failure modes, and architectural decisions.

50 Theory Questions7 Implementation Folios5 Free Model Answers
THEORY QUESTIONS & SOLUTIONSShowing 5 of 50 questions
Q1Detail the lifecycle of HotSpot JVM thread synchronization locks (Lock Inflation).
  • Mark Word: An object's memory header contains a "Mark Word" storing locking state bits.
  • Stage 1: Biased Locking (Deprecated/Removed): Thread ID is written into the Mark Word. Subsequent locking attempts by the same thread skip synchronization operations completely.
  • Stage 2: Lightweight Locking: If another thread attempts acquisition, the lock inflates to lightweight. The JVM uses CAS to copy the Mark Word to the lock-seeking thread's stack. If it succeeds, the lock is acquired. If contention occurs, the thread spins briefly.
  • Stage 3: Heavyweight Locking (Inflated): If contention persists, the lock inflates to heavyweight. The Mark Word is rewritten to point to an OS-level monitor (ObjectMonitor). Threads are parked and queued in native OS wait states, incurring heavy kernel context-switching penalties.
Q2Compare JVM Garbage Collection algorithms: Serial, Parallel, G1, ZGC, and Shenandoah.
  • Serial GC: Single-threaded; pauses all application threads ("Stop-the-World"/STW) for both young and old collections. Ideal for tiny memory footprints.
  • Parallel GC: Multi-threaded young and old collection. High throughput, but causes significant STW pauses.
  • G1 (Garbage-First) GC: Divides the heap into equal region blocks. It targets regions containing the most garbage first, performing incremental, concurrent collections to meet configurable maximum pause times (-XX:MaxGCPauseMillis).
  • ZGC: A scalable, low-latency concurrent collector using colored pointers (storing metadata in reference bits) and load barriers (intercepting references to update object coordinates on-the-fly). STW pauses are sub-millisecond, independent of heap scale (supports terabytes).
  • Shenandoah: Similar to ZGC, uses concurrent evacuation barriers to compact the heap concurrently, minimizing pause times.
Q3Explain memory barriers and compilation instruction reordering under the JMM.
  • Instruction Reordering: Compilers, JVM, and CPU architectures reorder execution instructions to maximize pipelining and cache efficiency, provided single-thread execution outcomes remain unchanged.
  • Memory Barriers (Fences): Hard CPU instructions inserted by the JVM to enforce memory ordering:
    • LoadLoad: Prevents subsequent reads from reordering before previous reads.
    • StoreStore: Prevents subsequent writes from reordering before previous writes.
    • LoadStore: Prevents subsequent writes from reordering before previous reads.
    • StoreLoad: The strongest fence; forces all previous writes to main memory before subsequent reads are evaluated.
Q4Detail Classloading resolution, initializing sequences, and Class.forName() vs ClassLoader.loadClass().
  • Execution order:
    1. Loading: Reads bytecode binary streams.
    2. Linking: Verifies bytecode structure, prepares static fields with default values, and optionally resolves symbolic references.
    3. Initialization: Executes static initializer blocks (static {}) and assigns real values to static fields.
  • Class.forName("MyClass"): Loads, links, and initializes the class, running static blocks immediately.
  • ClassLoader.loadClass("MyClass"): Only loads the class; linking and initialization are delayed until the class is instantiated or referenced.
Q5How do you trace and diagnose memory leaks, thread starvation, and CPU spikes?
  • Memory Leaks: Generate heap dumps using jmap -dump:live,format=b,file=heap.hprof [pid]. Analyze references using Eclipse Memory Analyzer (MAT) to identify leak suspects retaining massive heaps (retained size).
  • Thread Starvation/Deadlocks: Generate thread dumps using jstack [pid] or jcmd [pid] Thread.print. Examine thread states to locate deadlocks or threads blocked indefinitely on monitor lock acquisitions.
  • CPU Spikes: Profile CPU hot paths using async-profiler or JProfiler to capture on-cpu stack traces, identifying performance-intensive methods and loop bottlenecks.

Unlock the remaining 52 Java (Hard) questions

You've completed the 5 free sample questions. Get unrestricted lifetime access to every question, model answer, implementation challenge, and all 27+ technologies for a single payment.

Unlock Everything — Already Enrolled? Sign In

₹399 India / $9 International · One-time settlement · Zero subscription

CROSS-DIFFICULTY NAVIGATION

Continue Preparing Java

EASY LEVEL

Foundations & Core Concepts

MEDIUM LEVEL

Mechanisms & Practical Patterns

Java Interview Questions (Hard) | DevPrep | DevPrep