I applied in-person. I interviewed at Persistent Systems
Interview
1.Written Test
CORE JAVA -30 hadoop -30 Spark 20
program :
Two files - take inner join on fileld ID using spark/map reduce
2. L1 technical
3. L2 technical
It will walk in and they called all at 9:30 but still up to 12 candidates were coming and giving test. They took 1 written test and who cleared test, sent to L1 and later to L2
Interview questions [1]
Question 1
L1 -Techincal
Inyroduce yourself
What is your project
What are your source data types ? -csv/RDBMS how you get it?
How bigger was the client cluster? What was data size?
Load was daily , weekly or month?
Why client selected hadoop rather than RDBMS?
Which tool for workfow?
What is staging in spark?
What is RDD?
What is intention behing lazy evaluation?
What is intention behind keeping RDD immutable/unable to update?
You have written multiple tranformations on your RDD but still you have not fired any action. How your spark server WEB UI will look like?
Suppose you fired action on on RDD what exactly happens internally in spark ?(Here I told about it goes backword 1 by 1 to created required RDD using lineage graph in backword direction and first RDD is calculated and again return back to action)
Which are the transformation in spark?
I have given you an RDD . how will you convert it to paired RDD uisng its first element as key? ans- RDD2=RDD1.map(lambda x:(x[1], x))
What is difference between hadoop 2X and 1X ?
What is HA concept? What if Name node failed? What to do and who was doing in your project?
What is heartbeats concept?
I have file 500 MB on hadoop 2x .how much block and replicas will be there ?
I have a file
home_id product meter
h1 p1 20
h1 p2 30
H2 p2 23
I want to create partitions with the key home id.How will do it on local file system without suing SPARK, HIVE ,MAP reduce. Use simle programing language like java/python.
Later how will you do it in hive and spark?
21. I have an 3x3 ARRAY which is sorted
1 3 5
7 8 9
11 15 18
Write a program so that if use passed any element from terminal, it will return its exact position in array. (i did as below )
a=int[3][3]
a=[(1,3,5),(7,8,9),(11,15,18)]
x=int(std.input()) --user input
For i in 1 to 3
For j 1 to 3
If x ==a[i][j]
Then print(‘location of x in %i %j’,i,j)
L2 : technical
1.there is file
Name id
Ajay 1
Ram 2
Ajay 3
Ram 4
Jack 6
Devid 7
ID is unique and Name might be repeatble.
Write program so that user will enter name ‘ajay’ then program will return list of IDs -[1,3]
Input Ram : output [2,4]
