Reporting AWS Tools RDS or Redshift?

My app is reporting stuff like dashboards and filters, based on filters forecast/sales etc data have to be shown. We build the UI with Angular or a similar UI. This is going to be in AWS cloud.

What tools should we be using for the back end?

Our application is read-heavy (hardly any updates/deletes) and every read has to perform analytics on the data. The set of tables is not more than 10-15 and each table may have at most 3 million rows.

We are deciding between RDS (any Oracle/SQL) and RedShift.

With the size of data that we have (it may grow in future but not exceptionally) is Redshift is going to be overkill?

We anticipate heavy analytics (calculations/aggregations etc).

We may have to load the data from S3 data files, possibly with Redshift Spectrum.

We plan to use AWS Lambda as middle-ware to fetch data for the UI.

In general, if an Amazon RDS database is sufficient for your needs, then it is preferable over using Amazon Redshift. Therefore, start with RDS and see how it goes!

More specifically:

• Amazon RDS is better for transactions (updating a small number of items in a database)
• Amazon Redshift is better for tables with millions and billions of rows (for reporting queries, not so much for small updates)
• If you're going to use Amazon RDS, consider using Amazon Aurora since it has excellent performance, reliability, scaling, etc
• Use an open-source engine if possible (MySQL, PostgreSQL, MariaDB) rather than a proprietary engine to lower costs and reduce lock-in