preface

Have you ever gone to an interview and been asked Flink’s interview questions that you couldn’t answer? Why was that? The food? It isn’t. The reason is that you are exposed to too few interview questions, so I will share with you today according to different groups.

1 Flink Foundation (suitable for beginners)

  1. A brief introduction to Flink
  2. How does Flink compare to traditional Spark Streaming?
  3. What are Flink’s stack of components?
  4. Does Flink have to rely on Hadoop components to run?
  5. How big is your Flink cluster?
  6. Do you know the basic programming model of Flink?
  7. What are the roles in the Flink cluster? What is the role of each?
  8. Describe the concept of Task Slot in Flink resource management
  9. What are the common operators of Flink?
  10. Tell me what you know about the Flink zoning strategy.
  11. Is the parallelism of Flink understood? What is Flink’s parallelism setting?
  12. What is the difference between Flink Slot and Parallelism?
  13. Does Flink have a reboot strategy? What are some of them?
  14. Ever used distributed caching in Flink? How to use it?
  15. What are the broadcast variables in Flink?
  16. What about Windows in Flink?
  17. What about state storage in Flink?
  18. What are the types of time in Flink
  19. What is the concept of watermark in Flink and what role does it play?
  20. Flink Table & SQL familiar? What does the TableEnvironment class do
  21. How does Flink SQL work? How is SQL parsing implemented?

2 Flink Intermediate (suitable for 1~2 years development experience)

  1. How does Flink support batch streaming in one?
  2. How does Flink achieve efficient data exchange?
  3. How does Flink do fault tolerance?
  4. What is the principle of Flink distributed snapshot?
  5. How does Flink guarantee Exactly-once semantics?
  6. What’s so special about Flink’s Kafka connector?
  7. How does Flink’s memory management work?
  8. How does Flink serialize?
  9. Window in Flink has data skew, what’s your solution?
  10. How to solve the data hot spots in Flink when using aggregation functions such as GroupBy, Distinct and KeyBy?
  11. Flink has high task latency. How do you solve this problem?
  12. How does Flink handle back pressure?
  13. How is Flink’s backpressure different from Strom’s?
  14. Do you know Operator Chains?
  15. When does Flink combine Operator chains to form Operator chains?
  16. What’s new in Flink1.9?
  17. How to deal with dirty Kafka data when consuming it?

3 Flink Advanced (suitable for more than 3 years)

  1. Flink Job submission process
  2. What are the “graphs” that Flink refers to as “three-layer graphs”?
  3. What role does JobManger play in the cluster?
  4. What role does JobManger play during cluster startup?
  5. What role does TaskManager play in a cluster?
  6. What role does TaskManager play in the cluster startup process?
  7. How is Flink computing resource scheduling implemented?
  8. Describe the data abstraction and data exchange process of Flink.
  9. How is the distributed snapshot mechanism implemented in Flink?
  10. How is FlinkSQL implemented?

4. Business Interview Questions (key points)

  1. Application architecture
  2. Pressure measurement and monitoring
  3. Why Flink when you have Spark
  4. Checkpoint of the storage
  5. The guarantee of exactly – once
  6. State mechanisms
  7. Massive key deduplication
  8. Compare checkpoint with Spark
  9. Watermark mechanism
  10. How exactly-once is implemented
  11. CEP
  12. Three time semantics
  13. Processing of data peaks

summary

Ok, that’s all for today’s Flink questions. Which questions are you afraid of being asked by an interviewer? Believe in yourself, hard work and sweat will always pay off. I’m big data, and I’ll see you next time

Answer: github.com/lhh2002/Fra…

Sweep yards attention