What I Learned From My Java Backend Interview: Collections, Thread Pools, Redis & AOP Insights

Preface

The author records a job‑search experience after a year and a half of work, aiming for a salary of 10K‑12K, reflecting on the challenges of finding a new position, the importance of searching while employed, and personal motivation.

Interview Summary

2020‑06‑04 Shenzhen Yunmai Technology 10K‑20K Java Mid‑level Engineer

Status: the interview was cancelled after a brief written test (84 points) and a rushed first interview without self‑introduction.

1. Project and Responsibilities

The candidate worked at a data‑security company on two projects: a database vulnerability scanning system and a big‑data AI‑driven database security governance platform. Responsibilities included data collection, ingestion into MySQL and Elasticsearch via Kafka, Redis caching for high‑frequency statistics, and managing database objects.

2. Collections Used

Frequent use of ArrayList, LinkedList, HashMap, and occasional Set.

3. HashMap vs Hashtable

HashMap

is not thread‑safe, faster, and allows null keys/values; Hashtable is thread‑safe, slower, and disallows null entries.

4. synchronized and volatile

synchronized

provides mutual exclusion for methods, blocks, or classes (object lock vs class lock). volatile guarantees memory visibility of a variable across threads.

5. Singleton Pattern

The singleton ensures a single instance, typically implemented with a public static field, a private constructor, and a public accessor method; eager (hungry) and lazy (lazy) initialization are the two main styles.

6. Singleton Thread‑Safety & Double‑Checked Locking

Without proper synchronization, lazy initialization can create multiple instances; double‑checked locking uses a synchronized block around the instance creation to guarantee only one object.

7. Java Thread Pool Types

Four common factories: newFixedThreadPool (fixed size), newCachedThreadPool (dynamic), newScheduledThreadPool (timed tasks), and newSingleThreadExecutor (single thread).

8. Spring Boot AOP

AOP inserts behavior before or after the original method execution without changing the core logic, commonly used for logging and transaction handling.

9. Spring Container Knowledge

The candidate mentioned bean lifecycle steps: component scanning → BeanFactory → BeanDefinition → injection and proxy creation, but could not elaborate further.

10. Other Spring Features

Unclear answer; the candidate recognized Spring’s broader ecosystem but did not specify additional functionalities.

11. Redis Usage, Cache Penetration & Eviction

Redis is used for fast data access rather than traditional caching. Cache penetration occurs when many requests query non‑existent keys, potentially overloading the database; solutions include proper key design and Bloom filters. Cache avalanche refers to massive key expiration causing a sudden DB load; mitigation involves staggering TTLs and distributed deployment. Common eviction policies (LRU, LFU, TTL, etc.) were outlined with diagrams.

12. Closing Remarks

The candidate expressed confidence in learning ability, willingness to improve technical breadth, and a positive attitude toward future growth.