Note: The server is down for the second time

       Real warriors dare to face the bleak life, dare to face the dripping blood, real programmers dare to face the bugs of online programs, dare to face errors and solve problems in time, bringing real improvement and sublimation.

       Looking back on the previous few months, that is, 2019, it seems not too far from this record, but the impression is relatively deep, and it can be considered that the project was really smashed by my own hands. At that time, the sunny afternoon was sunny. Mingmei, you can read this article  https://blog.csdn.net/chajinglong/article/details/103201325

       Looking to the future, in 2020, April is a dark day. The day before, my colleagues worked hard and worked overtime until three or four in the morning (I had something that day, so I went back after work and made a murmur. , I always feel something big happened), I took a rest the next day, I woke up after I was online, then I got up at 8 o’clock and washed my face, staring at the computer, thinking that everything was fine when I went online, and the tests were tested, and then I was ready to do it first I ate breakfast for the first time, my mind was turning quickly, and then the problem appeared...

      The problem happened?

     The first online project crashed at 10:30, and then I was notified urgently. I hurried to find the operation and maintenance to restart the project, and started to locate the problem. They said that it depends on the ability of a person, and the ability to locate the problem, and then perform There are sticks, knives, and guns in various kungfu... all of them are different. It depends on whoever comes first to solve the problem.

      Idea 1: I looked at the log, the user login crashed, all logins failed, and there were a lot of problems. Then I thought about what happened before. When I remembered the previous test, when the user logged in with a PC, he asked to log in on the APP side (remarks) : Single sign-on, if you don’t understand, you can read this article https://blog.csdn.net/chajinglong/article/details/85066243 ), and then users can’t log in all the time. It must be because of this. The problem appeared before it went online, and it was fixed before. Because it was accidental, could it be caused by this? Then I looked at the master code, looked at the logic, and found that updating two updates at the same time may cause deadlock. The reason seems to be found, I am happy now, resubmit, restart, and run.

      In the past one and a half hours, the problem occurred again. The online data rose sharply. The server and CPU were all normal. The user felt that the login failed. At first, we logged in and found that it was OK. After a few minutes, everything was OK. Can't log in,

To be continued...

 

 

 

 

Guess you like

Origin blog.csdn.net/chajinglong/article/details/105693678