0

I am addressing solution to a big data problem using AWS Redshift. This application now needs to be personalized which shall need storing user preferences in a data store. The users in the system are less than 100 and hence shall not require a big data store.

How shall I go about this, use a polyglot store may be a RDS Postgres to store user preference OR shall I use Redshift only to model user preferences.

ajjain
  • 1,151
  • 2
  • 15
  • 28

1 Answers1

1

It would be great if you could throw more light on your systems and use cases. In general Redshift is not that great for concurrency, and if your use case requires frequent reads to this user data, then probably you can keep that in any other transactional database.

But if your system is Analytical and you have your big tables in Redshift, and you also need smaller table to perform joins; you can very well keep that table in Redshift, just ensure that that Table is available on all Nodes(DISTSTYLE ALL). It will prevent from data being broadcasted and improve query performance

Rakesh Singh
  • 160
  • 4