Byte beating interview summary (August 2019)

Personal profile

I coordinate Nanjing, graduated 16 years, in a traditional industry in JavaWeb back-end development, three years of work experience. For technical anxiety, the ability to grow into a bottleneck, culture and atmosphere within the system unable to adapt to other reasons, determined to change the environment and work, access to the Internet industry, to continue his pursuit of the technology of the road. August 19 bytes candidates beating the company's back-end development engineer jobs, now for four technical experience and an HR surface to be summed up.

Interview Experience

Long time round technical 40min ~ 1h ranging from a possible round of interviews may also be two continuously. First round opening with a simple self-introduction, then the interviewer ask questions based on personal experience and technical ability, and finally set aside time for a 20min ~ 30min algorithm problem be Coding. According to the draft and memories, sorting out the incomplete four technical topics are as follows.

Interview questions

The first round of the second round &

1. Spring transaction propagation mechanism
2. Spring's annotation, IoC, AOP-related issues, including: private methods can be enhanced AOP
3. Redis memory failure / recovery strategy
4. Whether there is a message MQ duplication of
   two algorithms Title:
   1) The user goes online, offline logs, statistics the maximum number of concurrent
   2) Construction of a binary tree according to the preamble sequence and the sequence

Third round

0. Asked whether understand Big Data (because of my previous work with this content have nothing in common, so honestly told there is no understanding, in turn, ask the interviewer would give up other aspects)
1. OS memory fragmentation, memory management scheme
2. OS file system, file access mechanism
3. Distributed cache-related issues, including Key distribution / mapping strategy, action at Key does not exist
4. Principle Bloom filter
5. A typical HTTP request process
6. API gateway (proxy) function
   algorithm topic:
   quickly determine whether there is a number from the number in the billions; and discusses performance optimization algorithms related content.
   Note: In the last part of the interviewer to ask questions, the interviewer gave me a suggestion, learn and understand large data relevant knowledge, do not fully understand the back-end development of mainstream / hot technology direction, even if they are not related to technology and work.

Fourth Round

1. Talk of the advantages of the Spring
2. Technology pain points encountered in work projects and countermeasures
3. Talk about understanding BlockingQueue, and then implement a simple BlockingQueue
4. How do I match winning percentage is 0.6, then the 5 Board 3 wins and 3 innings, 2 wins the game which way the election? If 5 innings wins, winning percentage is P3,3 3 innings, 2 wins winning percentage is then deduced P2 7 Bureau of 4 wins and winning percentage P4
5. How is the work of learning and technology accumulation

HR surface

HR surface lasts about 30min, mainly we asked questions of the replacement work causes and ideas, current salary status.

Technical summary

Byte beating interview technique using bovine passenger network online interview platform (Interview link Example: https://people.toutiaocloud.com/recruitment/interview/interviewee/27d0d9c0-aeb5-11e9-a933-98039ba25b3e ), with the interviewer by video face to face, code written in the interview platform development environment, shared editorial content in real time to the interviewer.
Four technical experience technical interview, feel byte beating the technical requirements of candidates is relatively high, the candidates first need to master the technology stack work-related personal experiences of programming languages, Web frameworks, secondly to algorithms, data structures, operating system, network theory and other expertise to have a solid foundation, in particular, the ability of the algorithm to keep the title of hands Coding.

Experience

First talk handwritten code. Several years of work experience programmers tend to increasingly unfamiliar to the basic knowledge of algorithms and data structures, not to mention the usual work mostly business-related content, if not through deliberate training hard to fast answers about 20min time and achieve interview algorithm in question. In addition algorithm interview questions in terms of algorithms medium difficulty on the basis LeetCode entitled and standards, from withholding online byte beating algorithms in paper , the title of the current 335, 111 simple class title, medium difficulty of the subject 170, difficult topics 54.

I use his spare time about a month, using the deduction from net brush questions (about 60 medium difficulty topics) on learning and "classic data structures and algorithms to resolve the problem" (Java description language) way of combining a book, knowledge of data structures and algorithms to pick up again, the training capacity for Coding algorithms interview questions and be familiar with the process. Personal feeling brush only problem is that the state and the rhythm familiar with the test, the key lies in a solid grasp knowledge of data structures and algorithms, of course, provided that routine work often write code, the code is not unfamiliar handwriting, otherwise you will need more practice It can be done online from withholding about 170 medium difficulty algorithm problem.
Learning and accumulation of knowledge about the basic, if in order to prepare for the interview, conducted utilitarian learning, then the back-end development technology stack for JavaWeb candidates, at least to learn and master the content of the following books, including the Java language foundation and advanced standard library , Java8 new features, concurrent multi-threaded programming, Spring framework, Redis cache component, Java virtual machine GC and other related knowledge.

About accumulation of knowledge and skills, or should be long-term learning, it is best to develop the habit of lifelong learning. More books, I can not do is use the time outside of work in a few months time to learn and master of all; thanks to continuous learning and accumulated bit by bit from the beginning of the work began three years ago, more than books the basic read at least twice, and the photograph could not cited bibliography.

In addition to deliberate learning and coding exercises, individual work usually used in conjunction with technical learning content, constantly trying to learn new tools, new technology into the project to work, such as the introduction of Lambda expressions and stream processing, the pursuit of clean and elegant writing Code, developed the Java programming specification for code refactoring and iterative software, using Redis components and so on. Only to learn Zhiyong, in order to accumulate experience in practice, at least be able to calmly answer in the interview, there is theoretical knowledge with practical experience. The existence of a defect is not related to the content of the work, when the interviewer asked whether understand large data relevant knowledge, can only be embarrassed to answer fully understand, but fortunately this is still quite beat bytes inclusive.

Report: Ali two technical

Byte before and after beating interview lasted about three weeks, during which also experienced two rounds of telephone technical Alibaba Flying Pig travel business sector.
The first round of interviews lasted 40min, the problems encountered are:

1. Database of optimistic and pessimistic locking
2. JVM's garbage collector, the type employed in the practical application
3. Multithreading, volatile keyword
4. Consistent hashing algorithm
5. Redis, cluster deployment
6. JDK8 compared to the previous version to make improvements
7. How to promote the implementation of the project team within the coding standard
8. Which Java diagnostic tools to troubleshoot problems usually use

The second round of telephone interviews lasted 30min, problems include a message queue; ask questions to spike as the background of the actual line, including one million concurrent amount of cache service capacity expansion; on Linux servers, given the access log of a 10G file, the statistics of the most visited of the 10 IP. As usual job does not involve high concurrency scenarios, using the Windows Server Deployment Services, the interviewer finally gave me the conclusion is due to living in traditional industries and the Internet out of touch more serious front-line combat, so a second round of interviews successfully hang up.