Processing: 24% (EMR, Spark, Hive, Lambda, Glue, ECS)

Question 1

Name the four data processing methods

Answer 1

a) batch, for processing of massive datasets at once
b) periodic, for unpredictable workloads
c) near real-time, for small bursts of data that must be collected and processed within minutes
d) real-time, for tiny bursts of data that must be processed continually

Question 2

Name the four Hadoop modules

Answer 2

a) Common (or Core)
b) the Hadoop Distributed File System (or HDFS)
c) Yet Another Resource Negotiator (or YARN)
d) MapReduce

Question 3

Name the difference in purpose between Hive and Presto

Answer 3

Hive is optimised for query throughput whereas Presto is optimised for interactivity