在阅读 LEC01 讲义原文 前，请先思考以下问题：
- What is a distributed system?
- Why distributed?
- What will likely limit the performance?
- How does detailed design reduce effect of slow network?
- How do they get good load balance?
- What about fault tolerance?
- Details of worker crash recovery?
- Other failures/problems?
- What if the master gives two workers the same Map() task?
- What if the master gives two workers the same Reduce() task?
- What if a single worker is very slow -- a "straggler"?
- What if a worker computes incorrect output, due to broken h/w or s/w?
- What if the master crashes?
- For what applications doesn't MapReduce work well?
- How might a real-world web company use MapReduce?