1. Need guidance for our scenario. I have a table(SUBSCRIPTION) with 5 columns
(CID, SID, ELEMENT_NAME, VALUE, SDATE) with values below
1001 12341 NAME JOHN ABRAHAM 10.22.2015
1001 12342 ADDRESS GRAHAM BLVD 240 CHICAGO, 10111 10.22.2015
1001 12343 SSN 123 53 5252 10.22.2015
1001 12344 CONTACT 425 432 7643 10.22.2015
I understand that i could apply only one masking rule to a column name and i cannot apply it dynamically. However my scenario here is to mask the VALUE column data based on ELEMENT_NAME column data. Eg. For the customer 1001 i have to apply name haslov masking rule if I see NAME in previous column and USaddress haslov masking rule if i see ADDRESS in previous column data similarly apply USSSN masking rule if I see SSN text in previous column.
Is there a way I can dynamically apply masking rules to one column referring to a value present in anther column.
2. I have another scenario to mask pii element present inside comment column data. eg. For SUBSCRIPTION table has column names CID, SID, REMARKS, DATE with values below
1001 321412 'called customer John Abraham with his contact 4254327643. he replied to send post to his address at Graham blvd 240 chicago,10111' 09.15.2017
1001 322411 'called again and verified his ssn 123 53 5252 but it is incorrect, updated his social as 425 524 2223' 09.16.2017
1001 334124 'send post to his residence address Graham blvd 240 chicago, 1011 and sent text message' 09.19.2017
Maybe you can use where condition to split your masking. You'll mask first all data where ELEMENT_NAME = 'NAME', applying the required masking rules for this scenario. Then you mask all data where ELEMENT_NAME = 'USSSN', and so on.
Thank you Artur Santana, it helps for my scenario1. Guess, to have performance tuning scripts here. Will come back. Any help for the scenario 2.
The second scenario is more complex.
Do You really need all those comments, with original text, on a 'non-production' database?
Because you could create a seedlist with a list of generic comments and use it to mask this column, with RANDLOV.
If this is not an option maybe you can use PARTMASK with option N for numerics enable. FDM will mask, with random numbers, only the numeric part of the comments, masking the SSNs, phones and ZipCodes.
Yes. The comment data is extracted by other down stream applications for its own functionalities/operations processing. And so consistent masked data is required to maintain referential integrity.
Since the comments can be written in any form it will be really difficult to mask only portions of it. You would either need to mask the entire column or use PARTMASK to mask the characters or numbers.
Anil, I understand. PARTMASK is difficult because i do not know the exact PII element present in the comment.. it can be either phonenumber or name or address.. i guess the best is to nullify or make a fixed value. Thanks.
Yes that would be the best way forward.
Did this resolve your issue?