1, you write all the items, say it again
2, MR execution flow
3, flume fan-in fan-out
4, kafka the topic partitions have anything to do design and consumer
Each group can have multiple consumer, each consumer belong to a consumer group;
Under normal circumstances, a consumer group will contain more, so not only can improve concurrency spending power topic of messages, but also to improve the "fault tolerance "sex, if a consumer group in the partitions fail then the consumer will have other consumer automatically takes over.
The Topic for a particular message, in which a consumer will only be subscribed for each consumer group in this Topic This message will not be sent to a group of more consumer;
then a group all the consumer will be staggered consumer entire Topic, each group in consumer news consumption independent of each other, we can say that a group is a "subscription" person.
In kafka in a partition of the message will only be a consumer of the consumer group (the same time);
a Topic for each partions, a consumer will only be a consumer "subscribers" in, but a consumer can while consumption multiple partitions in a message.
Design Principles kafka decision for a topic, the same group can not have more than the number of partitions of the consumer at the same time consumption, otherwise it will mean that certain consumer will not get the message.
kafka can only be guaranteed when a partition in the message is a consumer consumption is sequential; in fact, from the point of view Topic, when there are multiple partitions, the message is still the global order.
5, tableau how to use.
6. Why hbase storage
7, hive optimization
https://www.cnblogs.com/swordfall/p/11037539.html
8, hive and data processing INCLINATION
https://blog.csdn.net/XiaodunLP/article/details/87866350
https://blog.csdn.net/jin6872115/article/details/79878391
Data inclined spark
https://blog.csdn.net/Fenggms/article/details/83020683
9, flink understand it
Understand some, Flink core is a streaming data flow execution engine, similar to sparkStreaming, often said to be the third generation of the stream data processing engine. But not used in the work, there are some of their own learning. There are two major categories API, for Streaming data type applications, providing DataStream API; for batch type applications, providing DataSet API (support for Java / Scala)
10, which used the data warehouse model, say a few, such as snow and star
11, which algorithms are used
12, java thread how to ensure synchronization
Multi-threaded above reference content
13, thread lock There are several ways to achieve
14, java GC garbage talk about
https://blog.csdn.net/laomo_bible/article/details/83112622
JDK8 Default Type GC: default is ParallelGC, corresponds to the collector:
Use of the new generation are: Parallel Scavenge collector
old's use is: Serial Old collectors,
JDK8 collector new G1
java -XX:+PrintCommandLineFlags -version
Key to answer contents: JVM composition (heap memory, stack), GC type, collector type, GC process, GC algorithm (detailed reference links)
15, jvm underlying how to play
16, a handwriting java design patterns
23 design patterns, such as the common single-mode embodiment, the factory mode, decorative patterns,
Overall, the design patterns into three categories:
1. Create type
Five kinds of: factory method pattern, abstract factory mode, single-mode embodiment, the builder pattern, the prototype model.
2. Structural
Total seven kinds: adapter mode, decorative mode, proxy mode, the appearance mode, bridge mode, combination mode, Flyweight.
3. Behavioral
A total of eleven kinds: Strategy pattern, the template method pattern, observer mode, iterator pattern, the responsibility chain mode, command mode, the memo mode state mode, the visitor pattern, intermediary model to explain the mode.
17, python cooked it, wrote a project Mody
Web Project to write Django project, reptiles project Scrapy project