Barclays Abinitio Developer Interview Question -

1) SCD  type 2 implementation, full load, daily delta load explanation?

2) What are the components that we are using in SCD type 2 implementation?

3) Types of parallelism?

4) Performance improvement technique?

5) Which component breaks pipeline parallelism?

6) Does sort within-group component breaks pipeline parallelism?

7) What all are the components you have worked on?

8) scenario - max of the transaction? if the data volume is high what should be the optimized approach?

i/p- partition by round-robin - Rollup ( 2 stage routing)

9) Count the number of records from a file in Abinitio? 

I/P - Rollup -O/P

Rollup - 

key - null 

transformation - use count function

10) What all are the script you have worked on?

11) m_db command, air command to check-in, checkout and compare with eme version copy?

12) Migration of code from one env to another environment?

13) Type of lookup file - Lookup and lookup_local?

14) In which scenario we should use lookup and Join?

14) (Unix) Remove empty lines from a flat file? 

15) Delete duplicate from a file without using Dedup sorted component?

Using Rollup key change or  Rollup in the expanded mode we can avoid duplicate(first and last function)

Using rollup in expanded mode -

 

I/P -

Transaction 

1500

1500

2000

2000

2000

4000

4000

Rollup key - Transaction

Transformation - out.Transaction :: first(in.Transaction);

or 

 out.Transaction :: last(in.Transaction);

Using Rollup Key change function - 

out::rollup(in)=

begin

out.*::in.*;

end;

out::key_change(in1,in2)=

begin

out::in2.Transaction !=in1.Transaction;

end;

15)Read multiple files and write multiple files?

16) (Unix) 30 days older files in a directory find command?

17) (Unix) Command to find the transaction not in 1500 and 2000 in a file?

grep -v '1500/|2000' input_file.dat

18) (Unix) Command to find the transaction greater than 2600?

awk '$1 > 2600' input_file.dat