How to mask PII data using DMS without adding a column

Question

I am replicating data from source db to target db within two different AWS accounts. I want to mask and replicate the column data to target db during ongoing replication in same column which contains PII data. I am using transformation rules for this but unable to do it. Because I don't want to add new column and delete the old one which contains PII data. 
Can anyone suggest me how to achieve this?

Zonix · Answer

Hi Abhijeet,

I'm encountering a similar issue with AWS DMS (Database Migration Service). There are numerous articles about data masking using DMS, yet they all fail to mention that actual data transformation within columns is not feasible. The transformations are only applicable to column names, which raises the question of whether modifying or deleting columns truly constitutes data anonymization.

My specific challenge involves an email column. I intended to remove this column and add a new one with standardized email addresses like name.surname@example.com. I followed this guide: AWS DMS Customizing Tasks.

I attempted to set this add-column rule through the DMS console using the following JSON code:
{
  "value": "FULL_NAME",
  "expression": "$FIRST_NAME||'_'||$LAST_NAME",
  "data-type": {
    "type": "string",
    "length": 50
  }
}
However, I encountered an issue because the expression field rules cannot be executed via the console. I also tried defining the transformation rules in JSON, but errors occurred. Consequently, I am unsure if these rules are actually implementable through the console.

Didier_Durand · Answer

Hi,

this blog post describes a solution to your need: https://aws.amazon.com/blogs/database/data-masking-using-aws-dms/

Best,

Didier

How to mask PII data using DMS without adding a column

Relevant content