how does MapReduce work levi87@levi87?

master divide job and assign to slaves, slaves work (and delegate to more nodes) and return results to master

the work count example may help~

that just one request tho?

1 master 1 map? or 1 master many maps

i'm picturing this hierarchically.. so 1 master do 1 map (delegate to 5 work nodes for example).. im not sure tho

in the example it seems like multiple maps go to the same reducer..

and it seems like the reducer is joining results from multiple maps?

ah i see wut u mean, i was talking about the mapping function per input XD"

The map function is applied in parallel to every item in the input dataset

--> list of (k2,v2) pairs for each request, which are later collected/grouped, and the groups passed into Reduce function

v_vlll: thanks la

anyone up for study group tomorrow?

think i will be cramming math tmr, and the exam doesnt end until late at night ><"

*pat pat* + oil!

I need to get out of the house and study T_T

i'm up for a study group, lol

sure, when/where?

please be someplace with easy internet access...

errm

waves cafe near patterson like that time?

don't ditch me again

facepalm, i dun know where that is

maps.google.ca/maps?sour...

free wi-fi, i think

as early as possible damn it

i barely started studying orz

10am @ waves

f*ck you DK XD

*puts on glasses* YEAHHHHHHHHHHHHHHHHh