0

I am reading about multi-region architecture considerations.

Our reasons for moving to a multi-region architecture are pretty much the same as everyone else's:

  1. Reducing latency for customers that are in different continents (EU, US, Asia, Africa)

  2. Being in compliance with their data storage needs

  3. Enable regional failover

Here: https://onica.com/blog/security/aws-multi-region-architecture/, It says

Reason #4: There are laws governing my data that mandate regional PII data must remain within that region. This is another scenario in which multi-regional architectures are the norm. Investigate an “active/active” architecture.

We will be using Cognito pools and dynamo DB for data storage. My understanding is that active-active needs me to replicate data in other regions(Global tables in DynamODB) but the constraint of keeping PII data in the same region as the customers suggest otherwise.

How should this be accomplished keeping in mind that the customers are spread across continents i.e Asia, US, EU e.t.c (so If I was to use global tables and replicate data in multiple regions, how should that be done in DynamoDB). Also, how should this be handled in cognito pools

systemdebt
  • 4,589
  • 10
  • 55
  • 116

1 Answers1

-1

For multi region there is no other way but to replicate the data. If you have a constraint of customer data not leaving the region, than your best option is to have separate identification system for each region. Probably your users don't need to log within multiple regions. AWSguarantees that region specific data won't leave the region unless you explicitly make it. Like with Dynamodb global

karjan
  • 936
  • 1
  • 7
  • 17