what does RR actually do?

Moderator: Concepts and Technologies for DS and BDP

Beiträge: 40
Registriert: 29. Jan 2017 09:26

what does RR actually do?

Beitrag von tz33quvu »

because Inputformat has read data as line list, why should we add rr in the structure?

My simple thought is rr makes(take word count as example) lines to [(word,1)] for mapper, so what the mapper should do is only groupbyKey, is that correct?

Thanks in advance

Beiträge: 12
Registriert: 18. Apr 2015 16:25

Re: what does RR actually do?

Beitrag von yc81reja »

I am not sure if you mean rr. Is that RecordReader?
If so, I wouldn't think mapper only does groupByKey().
Please refer to this
https://www.quora.com/What-is-the-purpo ... -in-Hadoop
And please note the table on page 33 of slide 4 that by reading text file the rr provides (offset, line) as input for mapper.


Zurück zu „Archiv“