Cache Consistency Strategies, Java Fundamentals, and Common Backend Interview Questions

Project Part: A list of typical interview questions covering project background, data flow, traffic analysis, anomaly detection, and cache‑database consistency (e.g., whether to update the database first or the cache first).

Fundamentals: Detailed explanations of HashMap implementation (bucket, linked list, red‑black tree), default capacity, thread‑unsafe scenarios, and how JDK 1.8 resolves them; differences between arrays and linked lists; common thread pool types (newCachedThreadPool, newFixedThreadPool, newSingleThreadExecutor, newScheduledThreadPool) and their parameters; volatile keyword characteristics; synchronization challenges without using synchronized/volatile; MySQL table design considerations; SQL WHERE clause tips.

Cache Consistency Discussion: Overview of why updating the cache after the database can cause dirty data, followed by three update strategies:

Update database then update cache.

Delete cache then update database.

Update database then delete cache.

Analysis of each strategy’s drawbacks, especially in concurrent scenarios, and introduction of the double‑delete (delayed) strategy with sample code:

public void write(String key, Object data) {
    redis.delKey(key);
    db.updateData(data);
    Thread.sleep(1000);
    redis.delKey(key);
}

Guidance on choosing the sleep duration based on read‑operation latency, handling read‑write splitting, and ensuring eventual consistency.

Failure Handling: When the second cache deletion fails, propose two retry mechanisms: (1) send the key to a message queue and repeatedly attempt deletion; (2) use a binlog subscription (e.g., Canal) to capture database changes and trigger cache deletion, with optional asynchronous retry.

Additional Cache Issues: Descriptions and solutions for cache penetration (using placeholder values or Bloom filters), cache avalanche (staggered TTLs, distributed locks, high‑availability), and cache “no‑bottom‑hole” (reducing mget calls, command optimization, connection pooling).

Java Thread Pool Deep Dive: Explanation of each pool type, their creation limits, thread lifecycle, and when threads may die or be recreated.

Interview Coding Problem: Given an integer array nums and an integer k , determine if the array can be split into groups of k consecutive numbers. Includes three examples (two true, one false) and constraints. Sample solution skeleton:

1 <= nums.length <= 10^5
1 <= nums[i] <= 10^9
1 <= k <= nums.length

Overall, the article provides a comprehensive guide for backend engineers preparing for interviews, covering both theoretical concepts and practical code implementations.