Master Java Interview Essentials: OOP, Spring Beans, Redis, DB Indexes & More

Alibaba Cainiao (Java Phone Interview)

1. Introduce Java object‑oriented features

Object‑oriented programming abstracts real‑world entities as objects with fields and methods. Java’s three OOP features are encapsulation, inheritance and polymorphism.

Encapsulation : combines data and behavior, hides internal details.

Inheritance : allows subclasses to reuse parent class structure.

Polymorphism : enables different classes to respond to the same message, including compile‑time (overloading) and runtime (overriding) polymorphism.

2. Difference between overloading and overriding

Overloading

defines multiple methods with the same name but different parameter lists in the same class. Overriding lets a subclass provide a new implementation for a method declared in its superclass, keeping the same signature.

3. Underlying data structures of ArrayList and HashSet

ArrayList

uses a dynamic array; HashSet is based on a hash table (HashMap).

ArrayList : dynamic array, initial capacity 10, expands by ~1.5×, allows duplicates, preserves insertion order.

HashSet : backed by a HashMap, stores elements as keys, disallows duplicates, unordered; in JDK 8+ long buckets become red‑black trees.

4. Bean lifecycle in Spring

Creation : instantiate bean and inject properties.

Initialization :

Invoke Aware interfaces.

Execute BeanPostProcessor before‑initialization.

Call InitializingBean.afterPropertiesSet or custom init‑method.

Execute BeanPostProcessor after‑initialization.

Usage : bean is ready for use in the container.

Destruction : call DisposableBean.destroy or custom destroy‑method when the container shuts down.

5. Which fields can use a composite index (a,b,c) in the query select * from table where a=? and b>? and c=? ?

The index can be used for a and b. Field c cannot be used for index lookup because b uses a range condition, but c can be filtered by index‑pushdown.

6. Differences among process, thread and coroutine

Process : OS resource‑allocation unit, independent memory space, heavy context switch.

Thread : CPU scheduling unit within a process, shares memory, lighter context switch, requires synchronization.

Coroutine : User‑level lightweight thread, scheduler controlled by the program, minimal switch cost, but requires explicit management.

7. Common IPC mechanisms in Linux

Pipe (anonymous and named)

Message queue

Shared memory

Semaphore

Signal

Socket (TCP, UDP, Unix domain)

8. Six principles of design patterns

Single Responsibility

Open/Closed

Liskov Substitution

Dependency Inversion

Interface Segregation

Law of Demeter

9. Typical uses of Redis

Cache

Leaderboard (sorted set)

Distributed lock

Counter

Message queue (pub/sub)

10. Redis data types

Core types: String, Hash, List, Set, Zset. Newer types (since later versions): BitMap, HyperLogLog, GEO, Stream.

11. Differences between relational databases and NoSQL

Relational databases store structured data in tables and support ACID guarantees. NoSQL databases store data as documents, key‑value pairs, hashes, etc., follow BASE principles, and provide easier horizontal scalability.

12. How to implement QR‑code login

Server generates a unique QR‑code ID, stores status “unscanned” in Redis, and displays the QR code.

User scans with a logged‑in mobile app, which sends the ID and its token to the server.

Server validates the token and marks the ID as “confirmed”.

Web client polls Redis, sees the confirmed status, and proceeds.

Server issues a session or token to the web client, completing login.

13. Brain teaser: measure 15 minutes with two incense sticks

Light one stick at both ends and the other at one end.

When the first stick burns out (30 min), light the opposite end of the second stick.

The remaining burn time of the second stick is now 15 minutes.