Comprehensive Java Backend Interview Guide: Collections, JVM, OS, Networking, MySQL, Redis, and Algorithms

Java Basics

Explains the List, Set, and Map collection families, the differences between ArrayList and LinkedList, and details common implementations such as ArrayList, LinkedList, Vector, Stack, HashSet, LinkedHashSet, TreeSet, HashMap, LinkedHashMap, HashTable, TreeMap, and ConcurrentHashMap.

JVM

Describes why Java is cross‑platform: source code is compiled into .class bytecode that runs on the JVM, which abstracts hardware differences, enabling 一次编译到处运行 . It also notes that objects are usually allocated on the heap, but escape analysis can allocate non‑escaping objects on the stack, and lists relevant JVM flags (‑XX:+DoEscapeAnalysis, ‑XX:+EliminateAllocations).

Details garbage‑collection generations: the Young Generation (Eden and two Survivor spaces) and the Old Generation, with Minor GC occurring in the young generation and Full GC in the old generation, illustrated by diagrams.

Operating System

Compares threads and processes, covering resource allocation, overhead, stability, memory sharing, and their hierarchical relationship.

Network

Provides step‑by‑step explanations of the TCP three‑way handshake and four‑way termination processes.

MySQL

Lists the four ACID properties of transactions and explains how InnoDB ensures them using redo log (durability), undo log (atomicity), MVCC or locks (isolation), and the combination of these mechanisms (consistency).

Illustrates concurrency phenomena—dirty read, non‑repeatable read, and phantom read—with concrete examples, and shows how the REPEATABLE READ isolation level mitigates phantom reads via MVCC and next‑key locks.

Mentions considerations when adding a new column to a table, specifically MDL lock conflicts between reads and writes.

Redis

Discusses cache consistency strategies: read‑through bypass cache, write‑through update‑then‑invalidate, and the impact of the CAP theorem (cache as AP). It presents two approaches for achieving eventual consistency: a message‑queue retry mechanism and binlog subscription using Canal, with diagrams of Canal’s workflow.

Algorithm

Presents a simple algorithmic exercise: finding the second largest number in a collection.