Naive implement of mit6.824 lab. Please don't use these code other than for reference.
- lab 1 √
- lab 2 √
Just implement what in the paper
Most part is directly derived from paper. Generally, i build it in event loop fashion.
And also some subtle details, e.g., the optimization mentioned from the bottom of page 7 to top of page 8,
which is said unneccessary in real production environment, is proven critical to pass test Figure8Unreliable2C.
So i figure it out with my own solution. You can found these details and my solution of them in the code.
What lacks: now, my implement can run 2B as fast as the given example on the lab's page. However, it takes 30s more to finish 2C more than the example does. I guess the read/write lock used in leader's work is the cause. I'll think about it later.