raft实现阅读

[runlu]

raft实现指南
https://zhuanlan.zhihu.com/p/26506491
详细的lab实现-基于go语言
https://pdos.csail.mit.edu/6.824/labs/lab-raft.html
raft官网
https://raft.github.io/
图形化了解raft（非常好理解）
http://thesecretlivesofdata.com/raft/
raft paper
https://raft.github.io/raft.pdf
raft中文
https://github.com/maemual/raft-zh_cn/blob/master/raft-zh_cn.md

[runlu]

https://www.cnblogs.com/liuhao/p/5179588.html
抛些问题出来，真正解释了这些问题才算理解了论文。:)

1. 什么是复制状态机

2. Raft vs Paxos

3. Raft的设计目标understandability，为达到设计目标在做设计时如何权衡

4. 有稳定leader时的行为：
   a) 如何确保不同副本之间的数据一致；
   b) 一条日志何时可以在状态机上执行；

5. 如何选举？
   在Leader切换时必须避免的： a) 双主(split brain) b) 长时间无主(选主是否可收敛) c) 丢失已确认日志 d) 主备数据不一致

6. split vote如何处理？

7. 如何选择超时时间？

8. 真的没可能同时存在两个Leader吗？如果有两个副本同时认为自己是Leader，对读写操作各会造成什么影响？

9. 对选举是否还有其他约束？为了达成什么目的。如果没有这个限制，bad case

10. 是否形成多数派的日志一定是已提交的日志？Raft的设计为什么会存在这个问题。

11. nop日志的作用，如果只考虑写操作，是否必须有nop日志？
    ############################################################################

12. Raft实现的性能瓶颈，如何解决？
    https://stackoverflow.com/questions/40078139/raft-how-to-solved-the-performance-bottleneck-of-leader-node
    In raft, all operation requests will be forwarded to the leader node, and then the leader will send logs to all followers. So under a heavy loaded environment, the leader node will be a bottleneck. How to solve this?
    A：
    This can be solved in different ways depending on your desires. Here are some example solutions.

Partition the data. Many large-scale systems partition the data to spread the load (as well as reduce hurt if a partition goes down). But transactions cannot cross partitions. That could be a bummer, depending on your application.

Chain Consensus. This protocol spreads the work of moving data to all the nodes in the cluster. There is still a leader that is a bottleneck for accepting data, but its burden is smaller. Chain consensus also leads to slightly higher latencies than a broadcast system.
http://dsrg.pdos.csail.mit.edu/2013/08/08/chain-replication/