-1

What are the best practices to keep the data QA/UAT to best represent all scenarios PROD environment?

The intention is to have the lower environment as close as PROD environment so that we can identify and test all scenarios in lower environment before deploying changes to Production.

One idea is to sync past X months data into UAT AND strip off / randomize / de-identify the personal identification information for privacy protection and data security.

Looking for suggestions, links to article or videos.

Kritul Rathod
  • 75
  • 1
  • 11

1 Answers1

1

let's say you have 1 table called prod-data. You can create another table with name uat-table. And use dynamodb-stream of first table and a lambda function to insert data in uat-table.

In the lambda function

a. you can remove the PII information

b. set ttl while inserting in uat-table

c. set lower concurrently of lambda function to limit number of wcu consumed.

d. set higher Batch size so that wcu can be less.

for more information read this documentation.

OR

you can use production table only giving access to only non pii data. read here.

PS this solution has lower cost but has multiple limitations.

best wishes
  • 5,789
  • 1
  • 34
  • 59
  • Thanks for your recommendations. IMO, Real time sync is not needed, a lambda trigger once a month should suffice to refresh the data. So instead of Dynamo DB Stream we could just query data during the scheduled Lambda Run and adjust the Batch size to minimize RCU / WCU consumption. – Kritul Rathod Apr 09 '20 at 12:43