# introduction
It has been more than two weeks since I got the offer of byte, and now the work at hand is almost finished. Today, I will summarize the process of byte interview, hoping to help those who want to go to Dachang.
Time to introduce
- The position is operation and maintenance research and development
- I received a call from Hangzhou from November 22 to 20.32 on Sunday to invite me for an interview at 11.24 20:00
- On Tuesday, 11-24, 18:00, hangzhou still called me and told me that the interviewer was temporarily engaged, and the interview was postponed to December 01, 20:00
- Beijing called me from November 27 to 16:08 on Friday and told me that the interviewer had something to do and the interview was postponed to December 03 to 20:00. Well, it’s a bit of a crash.
- Thursday 12-03 20:00~21:15 video side, using flying books
- HR told me that the first interview was approved at 13:30 on December 4, Friday, and the second interview will be held around 20:00 on December 6
- The interviewer will have an important meeting at 16:40 on December 6th, Sunday. The second interview will be postponed to December 7th, 20:00
- The second video interview will be conducted on Niuke from 20:00 to 21:15 on Monday 12th to 7th.
- I received a phone call from HR at 22:15 on December 7, 2007 that the second interview was approved, and the third interview will be held around 21:00 on December 15, 2015 (the span is a little long, because the interviewers are all booked in the evening, and I want to make adequate preparation).
- Tuesday 12-15 21:00-21:45 video three interviews, using niuke video interview.
- I received a call from HR from December to 17 at 19:50 on Thursday, the technical side was approved, and the HR interview will be held from December to 20 at 15:00
- Sunday 12-20 15:00~15:30+ HR meeting, using flying books
- Hr added me to wechat from 12-22 20:38 on Tuesday and told me to prepare materials such as salary flow
- Wednesday 12-23 12:20+ reply hr related materials
- Received the offer email from 12-25 19:04 on Friday, it was Christmas day.
One side
The main technical points are as follows
- Java memory model, runtime data area
- Synchronize Lock upgrade process and mechanism behind it (objects, etc.); Reasons for upgrading (pros and cons of spin locks and heavyweight locks)
- Garbage collector, collection algorithm, G1 collection process (card table, tricolor mark)
- Current limiting correlation algorithm, implementation, etc
- Cache three major problems, bloom filter implementation principle
- Load balancing implementation, mainly said common centralized NG, LVS, haProxy
- Thread pool knowledge, if let you design a thread pool what need to pay attention to
- Version of the HTTP protocol. 405, 502 Status code meaning
- TCP three times handshake, four times wave introduction, which also asked the timeout reason – network knowledge weak did not get to the point.
- Class loading mechanism, parent delegate model
- Springboot configures the file loading sequence
- How to implement a springboot like a scheduled task @schedule
- Spring loop dependency issues
- Common sorting algorithm, heap sort time complexity
- Algorithm – LeetCode 162. Look for peaks
Second interview
The two main technical points are as follows
Here are some of the interview questions.
- How is kafka data consistency guaranteed
- The use of Kafka in the project, other MQ introduction and comparison
- Redis persistence mode, the process of persistence
- Mysql related: index, MVCC, transaction isolation level, primary key index versus normal index which is fast
- Design a strongly consistent consumption model
- Es cluster introduction, primary node election process (this is related to my project experience)
- Load in Linux, where queue refers to what queue
- Java program performance problems investigation, mainly JDK own command, nothing can see again
- Arthas dynamic class substitution implementation principles, Instrumentation, ASM, etc
- Algorithm – LeetCode 198.
On three sides
List of major issues
- Why do you choose this position
- Message queue comparison if data consistency is achieved. The implementation principle behind RocketMQ transactions
- Redis high availability cluster related knowledge, free to play
- Contact with the monitoring system, introduced how to design a monitoring system, free play
- Cache breakdown, penetration, avalanche, free play
- No algorithm problem, probably because I can’t even write LeetCode Easy.
HR side
Personal introduction
What is the most successful project & introduction
Reasons for leaving previous company
Current salary and benefits, etc
Current role in the team
During the interview, it was confirmed that the mobile phone number could be added to wechat, and the follow-up contact was made through wechat
The last
Space is limited, the rest of the content will not be shown here one by one, from the three sides of Bytedance Java, hot side by sharing, need answers to the small friends can pay attention to the public account: Kylin to change the bug to receive. Welcome everyone to exchange, like the article remember to pay attention to me like yo, thank you for your support!