MDM Interview Questions
Content
Interview Questions
Sunil Interview question:
It's support a project Tier 2 and 3 support
1.
Tell us about your experience.
2.
Have you spent more time on unit testing to understand how the system
is working?
3.
When you run the match job, on which server is it running ? (Cleanse/Ma
tch server)
4.
What is the issue you face while match process? What is the reason for
this table space issue comes ?
Ans: Tablespace issue. Load process fails because of the tablespace issue. The t
emp files are created which are
5.
In the process for matching, the consolidation indicator is used. Let's
assume the consolidation indicator is 3 for 50 records and 4 for 40 records, 2
for 10, 9 for 10 records. How many records will participate in matching?
a.
50 records will take part in match.
6.
Have you done Match & Merge optimization? how?
2.
Initially we defined 37 match rules but it impacts performance. We star
ted fine tuning the rules and reduced it to 30 and still working on reducing it.
There were many loose rules and then we created a combination of fuzzy and exac
t match rules.
7.
Did you work on real time or batch processing ?
3.
Batch processing
8.
Do you have experience in debugging Java code?
9.
Have you worked on IDD?
10.
4.
11.
Does IDD provide any capability where user can create a query?
Yes, we can create
Have you worked in hierarchy manager?
12. When a user logs in to Hub console and gets an error and unable to launch
the application. Which logs files will you check to debug the error? (Ans: cmxse
rver.log)
13. Load job has failed. Which log file will you look at to debug? ( database
log file)
14.
Stage job has failed. Which log files will you look? (cmxserver.log)
15.
How do we enable the data base in the hub? (In the C_Repos table)
5.
16.
Understand the structure of the C_Repos table.
Which version of Informatica you worked on ? 9.1
17. What are the downstream systems you have worked on after MDM process compl
etes?
6.
OLAP systems for reporting and also loaded back to the source systems.
18. we are loading data into party. Party has an attribute called party type.
There is a lookup table for the party entity. From one of the source systems, th
e party type has no data. This party type is in lookup table as well. What could
be the issue?
7.
19.
We have to enable the null values.
How do we enable the foreign key null value flag?
Pratik Interview
1. What have you done?
2. What are hierarchies? How do you create and give a real time example ?
3. Source one had 100 records full refreshed and source 2 has 50 records increme
ntal? Previous landing table?
4. What are the integration techniques you used for ETL?
5. After staging, I don't want any data in mapping. How will you remove the data
?
6. What are the core components of MDM?
Indira Interview
1. What were the types of trust?
2. In Installation, before configuring the Oracle there is one more step?
3. What is user exit?
4. Do you know about Trillium?
5. What is a Base Object?
6. Why do you go for MDM? Why is it necessary?
7. What is master data?
8. What are hub state indicators?
9. What are consolidation indicator?
10. In the match process, did you use any match calculation or match scores? How
did you find out that the record is golden?
11. Is model necessary for Cognos reporting? What are the components of Cognos?
12. Informatica: asked about process?
ShivamNovanor Disk- Leading healthcare company. 90 yrs. Focused on diabetes care. Prod
uct: needles, syrenges, 70,000 employee in 180 countries.
CRM - OLTP - Real time - 2000 records/day
Legacy - OLTP - Real Time - 2000 records/day
ECG - OLAP - Batch - 2 million records/week
How many base objects? - 52
BO - party model - Party BO- Customer BO - Customer Address BO - Investigator BO
- Customer Electronic BO
Where are the Pre Landing Tables created?
Where do you enable Delta Detection?
Shivam Interview:
1. In match & merge, and it takes 5 hours. how will reduce this time?
ans: No match job runs for 5 hours. Match process should be an hour or less. Ma
tch rules tuning.
Potential hot spots- large number of matches properties- dynamic match analysis
threshold. Tune the rules. Performance will improve if you tighten the rules. I
ncrease ram, improve
2. What have you done in IDD?
3. What is MRM? Why do we use it?
Master reference manager.
4. If you are not able to find data in the Base Object? What will you do? Why th
is happens? Sometimes data does not come, then what do you do?
You need to see the load job summary. It will tell the error. Ask him what is th
e case, lot of cases data gets rejected. sometimes Lookup value is missing. The
correct
Lookup tables are used in the load process. Lookups can be defined between base
object and base object and also betwen BO and Staging BO.
5. If no data has come to Base Object(load) then what will you do?
What is Immutable rowid:
What is distinct rowid:
validation rules
key types
key width
match purpose
match level
batch groups
sequence of running a job.
security access manager.
metadata manager.
6. If you want to delete some data from BO or all data from BO, how will you do
that ? (Soft Delete or Hard Delete?)
You can delete from hub console. You can delete from the database. clearscript.
run the stored procedure.
We can delete data from data manager. Create a query-create package. then we can
option to update or delete.
7. If you delete data, then what is it called Hard delete or soft delete? Can yo
u soft delete data ?
Can you do soft delete manually?
8. If you delete data by mistake then how will you get it back?
No one is allowed to touch the data in the database. Are you asking if it is dur
ing the production or is it during the development. Just load it from the
Hard Delete Detection is enabled in BO. It's external thing in database.
9. After you get golden data, how do you publish? What is the publish process?
10. If you perform any process in MDM, how can you see this online?
Indira Interview Quest:
1. After you migrate to another environment and any changes have been made to OR
S, how to do you reflect the changes? (Change list)
2. When you add a column to the base object, do you automate the process or you
do it manually?
3. Similarly, after the load process another column is added to the BO? What wil
l happen to the previous process? Can you add column after load process is done?
how will the existing running processes be impacted?
4. What were you using IDD for?
5. How did you do cleansing? Informatica power center was used for cleansing?
6. What is volume of the model? What is the data volume? (40 BO)
7. How many data records were generated after merging?
8. Batch job? How do you execute the batch groups and how do you check the statu
s of batch groups? How do you schedule the batch group?
9. IDD: when you change the data in ORS how the data is updated in IDD? do you d
o it manually?
10. What were you using IDD for?
11. What type of cleansing functions you used?
12. If
3 Types of records come into Base Object:
1. Unique record: It will go thru exact match process and then Informatica MDM H
ub will change its consolidation indicator to 1 and send it to Base Object
Question: We apply match rule on columns but what about the record?
2. Exact duplicate records:
3. Similar records
Sunil Interview question:
It's support a project Tier 2 and 3 support
1.
Tell us about your experience.
2.
Have you spent more time on unit testing to understand how the system
is working?
3.
When you run the match job, on which server is it running ? (Cleanse/Ma
tch server)
4.
What is the issue you face while match process? What is the reason for
this table space issue comes ?
Ans: Tablespace issue. Load process fails because of the tablespace issue. The t
emp files are created which are
5.
In the process for matching, the consolidation indicator is used. Let's
assume the consolidation indicator is 3 for 50 records and 4 for 40 records, 2
for 10, 9 for 10 records. How many records will participate in matching?
a.
50 records will take part in match.
6.
Have you done Match & Merge optimization? how?
2.
Initially we defined 37 match rules but it impacts performance. We star
ted fine tuning the rules and reduced it to 30 and still working on reducing it.
There were many loose rules and then we created a combination of fuzzy and exac
t match rules.
7.
Did you work on real time or batch processing ?
3.
Batch processing
8.
Do you have experience in debugging Java code?
9.
Have you worked on IDD?
10.
4.
11.
Does IDD provide any capability where user can create a query?
Yes, we can create
Have you worked in hierarchy manager?
12. When a user logs in to Hub console and gets an error and unable to launch
the application. Which logs files will you check to debug the error? (Ans: cmxse
rver.log)
13. Load job has failed. Which log file will you look at to debug? ( database
log file)
14.
Stage job has failed. Which log files will you look? (cmxserver.log)
15.
How do we enable the data base in the hub? (In the C_Repos table)
5.
16.
Understand the structure of the C_Repos table.
Which version of Informatica you worked on ? 9.1
17. What are the downstream systems you have worked on after MDM process compl
etes?
6.
OLAP systems for reporting and also loaded back to the source systems.
18. we are loading data into party. Party has an attribute called party type.
There is a lookup table for the party entity. From one of the source systems, th
e party type has no data. This party type is in lookup table as well. What could
be the issue?
7.
19.
We have to enable the null values.
How do we enable the foreign key null value flag?
Pratik Interview
1. What have you done?
2. What are hierarchies? How do you create and give a real time example ?
3. Source one had 100 records full refreshed and source 2 has 50 records increme
ntal? Previous landing table?
4. What are the integration techniques you used for ETL?
5. After staging, I don't want any data in mapping. How will you remove the data
?
6. What are the core components of MDM?
Indira Interview
1. What were the types of trust?
2. In Installation, before configuring the Oracle there is one more step?
3. What is user exit?
4. Do you know about Trillium?
5. What is a Base Object?
6. Why do you go for MDM? Why is it necessary?
7. What is master data?
8. What are hub state indicators?
9. What are consolidation indicator?
10. In the match process, did you use any match calculation or match scores? How
did you find out that the record is golden?
11. Is model necessary for Cognos reporting? What are the components of Cognos?
12. Informatica: asked about process?
ShivamNovanor Disk- Leading healthcare company. 90 yrs. Focused on diabetes care. Prod
uct: needles, syrenges, 70,000 employee in 180 countries.
CRM - OLTP - Real time - 2000 records/day
Legacy - OLTP - Real Time - 2000 records/day
ECG - OLAP - Batch - 2 million records/week
How many base objects? - 52
BO - party model - Party BO- Customer BO - Customer Address BO - Investigator BO
- Customer Electronic BO
Where are the Pre Landing Tables created?
Where do you enable Delta Detection?
Shivam Interview:
1. In match & merge, and it takes 5 hours. how will reduce this time?
ans: No match job runs for 5 hours. Match process should be an hour or less. Ma
tch rules tuning.
Potential hot spots- large number of matches properties- dynamic match analysis
threshold. Tune the rules. Performance will improve if you tighten the rules. I
ncrease ram, improve
2. What have you done in IDD?
3. What is MRM? Why do we use it?
Master reference manager.
4. If you are not able to find data in the Base Object? What will you do? Why th
is happens? Sometimes data does not come, then what do you do?
You need to see the load job summary. It will tell the error. Ask him what is th
e case, lot of cases data gets rejected. sometimes Lookup value is missing. The
correct
Lookup tables are used in the load process. Lookups can be defined between base
object and base object and also betwen BO and Staging BO.
5. If no data has come to Base Object(load) then what will you do?
What is Immutable rowid:
What is distinct rowid:
validation rules
key types
key width
match purpose
match level
batch groups
sequence of running a job.
security access manager.
metadata manager.
6. If you want to delete some data from BO or all data from BO, how will you do
that ? (Soft Delete or Hard Delete?)
You can delete from hub console. You can delete from the database. clearscript.
run the stored procedure.
We can delete data from data manager. Create a query-create package. then we can
option to update or delete.
7. If you delete data, then what is it called Hard delete or soft delete? Can yo
u soft delete data ?
Can you do soft delete manually?
8. If you delete data by mistake then how will you get it back?
No one is allowed to touch the data in the database. Are you asking if it is dur
ing the production or is it during the development. Just load it from the
Hard Delete Detection is enabled in BO. It's external thing in database.
9. After you get golden data, how do you publish? What is the publish process?
10. If you perform any process in MDM, how can you see this online?
Indira Interview Quest:
1. After you migrate to another environment and any changes have been made to OR
S, how to do you reflect the changes? (Change list)
2. When you add a column to the base object, do you automate the process or you
do it manually?
3. Similarly, after the load process another column is added to the BO? What wil
l happen to the previous process? Can you add column after load process is done?
how will the existing running processes be impacted?
4. What were you using IDD for?
5. How did you do cleansing? Informatica power center was used for cleansing?
6. What is volume of the model? What is the data volume? (40 BO)
7. How many data records were generated after merging?
8. Batch job? How do you execute the batch groups and how do you check the statu
s of batch groups? How do you schedule the batch group?
9. IDD: when you change the data in ORS how the data is updated in IDD? do you d
o it manually?
10. What were you using IDD for?
11. What type of cleansing functions you used?
12. If
3 Types of records come into Base Object:
1. Unique record: It will go thru exact match process and then Informatica MDM H
ub will change its consolidation indicator to 1 and send it to Base Object
Question: We apply match rule on columns but what about the record?
2. Exact duplicate records:
3. Similar records
