10/16 Class

Question 1

OLAP

Answer 1

online analytical process
main goal: to support ad-hoc but complex queries
Places key performance indicators(measures) into context(dimensions):
measures are pre-aggregated
data retrieval is significantly faster

Question 2

OLTP

Answer 2

online transaction process

Question 3

Why OLAP

Answer 3

dimension modeling is a natural presentation of data for business analytics.
OLAP Technology is very fast
Most reports run within 1-3 seconds
Speed advantage substantial in highly aggregated reports such as multi-year trends
Without OLAP, the burden is on the developer to extract relevant data and build aggregations
Pre-calculated results
Produces consistent information

Question 4

Roll up(drill-up)

Answer 4

summarize data by climbing up a concept hierarchy or by reducing dimensions
example:drill up from city to state
drill up by reducing the location dimension

Question 5

Drill down

Answer 5

analyze more detailed data by moving down a concept hierarchy or by adding dimension(s)
example:
drill down from city to dealer
drill down by adding the time dimension

Question 6

slice

Answer 6

creates a slice form the cube by choosing a single value for one of the dimensions

Question 7

dice

Answer 7

creates sub cube from the cub by choosing a two or more values for one or more of the dimensions

Question 8

hadoop

Answer 8

big data warehouse
 volume
 velocity
 variety
hadoop - apache open source software for reliable, scalable, distributed computing

Question 9

mapreduce

Answer 9

programming model invented at google in 2012
read in input and produce a key value pair
example:
key value

map workers: certain computers are used for this
reduce worker: certain computers only do this

Question 10

exam review

Answer 10

understand the difference between bitmap and B+tree
ETL
policies for data warehouse maintenance is on maintenance slide, user driven policy vs warehouse driven policy