关于“Exploring Google Ngrams with Amazon EMR”的评价

I got thought this one in time, just. But that was because I cheated. I first did the lab in my own account and that took well over 2hs. That may also be because I did a select count(*), which didn't return after 30 mins. I cheated also by upping the cluster size and machine type. It was good to play about with Ganglia, a little disappointing to use Hadoop 1. I'd actually like to repeat the exersize again using spark to see what the difference in cost and time are.

Jon S. · 评论about 10 years之前

John D. · 评论about 10 years之前

Hmm.. not done yet

Winzter C. · 评论about 10 years之前

Certain availability zones (N.Virginia) do not have the required instances, so cluster was terminating automatically. Wasted some time trying to find the right availability zone.

Lesley T. · 评论about 10 years之前

Geeta C. · 评论about 10 years之前

Guilherme V. · 评论about 10 years之前

Geoffrey M. · 评论about 10 years之前

Chitresh S. · 评论about 10 years之前

Praveen S. · 评论about 10 years之前

Karen H. · 评论about 10 years之前

sujeet s. · 评论about 10 years之前

Ian W. · 评论about 10 years之前

Alfonso T. · 评论about 10 years之前

this is a great intro to mapreduce

Eric K. · 评论about 10 years之前

Viorel D. · 评论about 10 years之前

Jade S. · 评论about 10 years之前

Andromancer D. · 评论about 10 years之前

Renaud S. · 评论about 10 years之前

Damon T. · 评论about 10 years之前

Aldo O. · 评论about 10 years之前

rob b. · 评论about 10 years之前

Renan C. · 评论about 10 years之前

DJ N. · 评论about 10 years之前

Whew! Some heavy lifting here!

Jim T. · 评论about 10 years之前

Matthew H. · 评论about 10 years之前