Change of plan:
Today I want to read some influential system papers from the industries. Two papers are selected. One is about MapReduce: "J. Dean and S. Ghemawat, MapReduce: Simplified Data Processing on Large Clusters, OSDI'04". The other is about BigTable: "F. Chang and et al., Bigtable: A Distributed Storage System for Structured Data, OSDI'06". Further comments will be added at the end of the day.
Plan today:
1. Finish reading the textbook 725. Read the notes.
2. The code for global trees. Talk with Brian about this.
3. The annotation writing for 788 and composition.
4. Read ACL papers and the OSDI04, 06.
I did not have the time to read the Bigtable today. I will read it tomorrow.
The mapreduce is a interesting model about parallel computing. The concepts seemed easy to understand and use, but the amazing part to me is the fault tolerance, load balancing and task scheduling. I just read half of this paper, and I will continue read this and the critics about Mapreduce.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.