DistOS 2019F 2019-11-06
Jump to navigation
Jump to search
Readings
MapReduce & BigTable
- Dean & Ghemawat, "MapReduce: Simplified Data Processing on Large Clusters" (OSDI 2004)
- Chang et al., "BigTable: A Distributed Storage System for Structured Data" (OSDI 2006)
Discussion Questions
MapReduce
- What is a map operation? What is reduce? How do they relate to functional programming?
- What kind of tasks was mapreduce designed for?
- How does mapreduce scale? How does it deal with failing nodes?
- How does mapreduce deal with stragglers?
BigTable
- What problem(s) is bigtable designed for?
- Is BigTable a relational database?
- How is data structured in BigTable?
- How does BigTable relate to GFS?
Other
- Do you think MapReduce and BigTable are still used at Google? Why or why not?
- What if you aren't at Google? Can you use the same kinds of technology?
- Do you notice any design patterns in what Google built?
- Other thoughts and questions?