Original address:http://hbase.group/article/89 <http://hbase.group/article/89>

The following interview questions are provided by the small partners in the group, On the spot( Including school recruitment questions)
1. Interview questions of Netease big data

* Talking about project
* Spark Which part works well, How to tune
* Java Which part is better to know
* Chat and concurrently, Concurrent implementation method,volatile Key words
* HashMap The underlying principle of
* Why rewritehashcode andequals
* say somethingjvm
* What is the application of each garbage collector
* jvm tuning
* say somethingio
* Why do you think about changing careers? Is it because the original major is not good for employment?

2. Interview questions of ant financial

* Merge of small files
* MR AndSpark Difference
* Blog about celebrities
* What's your opinion on big data
* How to learn big data
* StringBuilder AndStringBuffer Difference
* HashMap AndHashtable Difference
* Talk about your understanding of trees
* The realization of database index
* jvm Memory model of
* jvm Garbage collector for
* jvm Garbage collection algorithm of
* HDFS Framework
* HDFS Read and write flow
* Hadoop3.0 What improvements have been made
* ChatYARN
* Why the project chooses to useSpark, You feelSpark What are the advantages of
* understandFlink AndStorm Well, They andSpark Streaming What's the difference
* 1TB file, Take repeated words,top5 Under the scene of the specified asset, How to make statistics quickly
3. Interview questions of Meitu

* Why choose beautiful pictures, Do you know where metu is
* Introduce your project
* Data unified management platform, I'm interested, Tell me about it.
* I probably know what's going on,java web Did you take part in the development
* You just mentioned metadata in the project, Can you talk about it?hive Metadata management of, Do you know about it
* stillhive, You are right.hive What is the principle understanding
* knowAST,operator tree What do these look like
* Then yourhive turnmr How is the process understood?
* Except for predicate push down, Can we talk about other optimizations? Let alone the tuning of data skew
* jvm Understand not, Talk about garbage collection algorithm
* Ping Yongjava andscala Which language is more
* If I want to use it nowmap aggregate, What kind of access do you think is suitable for multithreading
* How to monitor threads
* Spark AppearOOM, How do you think to tune it? Do not movejvm Parameters
* You feeljoin How to optimize
* What is your plan for the future?( Within five years)
* You just take the technical route

4.​ Two faces of beautiful pictures

* ThriftServer OfHA How to achieve, Can you talk about the way to realize it
* It's a deal.Zookeeper Ofwatch How does the mechanism work?
* Scenario questions:
  Now there are1 individualclient,2 individualserver, When I join a machine dynamically, Or delete a machine, Or a machine goes down,client How to perceive, How to realize( No useZookeeper), How to communicate, Talk about specific implementation?
5. Interview questions of qiniuyun

* Quick row
* hive andhdfs Connection between
* inode And file descriptors
* linux Instructions how to create a file
* http inheader Put inkey value What's the change
* Difference between system call and library function
* http Buffering for wit
* session cookie  Difference
* Interprocess communication mode
* jsp essence
* http Request status
* get post put remove
* data basejoin 
* database engine
* hibernate andmybiters Difference
* jvm Garbage collection
* hive Different from relational database
* hive Realization principle
* spark Andmr Difference