Midterm Review
CSE4/587
B.Ramamurthy
5/28/2016
B.Ramamurthy
1
Exam Date
October 22, 2010
Please bring

Pencils, pens and erasers.
This is a closed book exam.
You can bring 4 sheets of any
information you need.
5/28/2016
B.Ramamurthy
2
Topics
Defining data intensive computing
Relationship between data-intensive computing and
cloud computing
Lucene
Enabling Technologies (ET):
 ET1: Web services
 ET2: Virtualization
 ET3: Special data structures and algorithms
(MapReduce Engine, HDFS)
Google App Engine
Project 1
5/28/2016
B.Ramamurthy
3
Data-intensive computing
competencies
1. Describe data-intensive computing concepts
2. Compute with data-intensive computing concepts
3. Recognize a data-intensive problem.
4. Identify the scale of data.
5. Analyze the data requirements of a problem.
6. Describe the data layout and define the data repository format
7. Decide the algorithms (Ex: MapReduce)
8. Define application-specific algorithms and analytics
9. Design the data-intensive program solution and system configuration.
10. Implement the data-intensive solution and test the solution for
functional correctness and non-functional requirements.
11. Research algorithmic improvements and experiment with them.
12. Write a report summarizing the solution and results.
13. Incorporate services from cloud computing platforms.
14. Integrate semantic information in organizing data for contextawareness.
15. Apply collective intelligence methods for diverse data sources.
16. Formulate data-intensive visualization solutions for presenting the
results.
5/28/2016
B.Ramamurthy
4
Take home questions: Q1
Read the list of
competencies and
check whether you
have learned this
competency in your
project 1 and/or in
the course lessons.
Item# from
slide 4
5/28/2016
Covered
in Prj1
Covered
in lessons
B.Ramamurthy
Where?
How?
What
extent
(%)
5
Take home questions: Q2
In order handle the massive load we had in
out recent deployment of the Pop!World
application Google App Engine (GAE), we
deployed 20 instances of the application and
gave the users the 20 links and let them
choose their link that was available when
they it ran out of quota.
Design a front end (method) using GAE
resources or others that will (i) hide the 20
instances or n instances and (ii) direct the
user to the one of the unloaded app. (The
policy could even be a simple round robin).
5/28/2016
B.Ramamurthy
6
Format of in-class exam
4 questions
Enabling technologies: Webservices,
virtualization, Mapreduce, HDFS,
infrastructure management
Amazon cloud environment (read the
blue book)
5/28/2016
B.Ramamurthy
7
How to study?
5/28/2016
B.Ramamurthy
8