What is mapreduce in hadoop
- what is map
- what is reduce
- example with graph
mapreduce for incremental
incoop techniques
challenges: transparency challenges: efficiency
Inc-HDFS
Incremental Map (first run VS second/further runs)
Incremental Reduce
Questions:
diferent aditional phases plain hadoop memoazation and contration contration performs it stuff automatically/manually ?
what would be a better decision for this automatically lots of grained or less but bigger inputs (similar, esta nos results)
memoization, what is it ? what happens if we skip it ? prevent repeated computations (cached value) we will degrain to base regular mapreduce