OLAP - Online Analytical Processing :

  • Purpose: Analyse large volumes Stored Historical Data.
  • its optimized for Read-heavy operations across tables
  • mostly used for analytics and BI tools
  • example: clickhouse.
  • usually transactions happen in the OLTP db and then occasionally dumped into the OLAP db.(its evolving)
  • Modern pipelines use something called CDC (change Data capture) to stream changes in near real-time from OLTP → OLAP.(read about Write-Ahead Log for depth. - PostgreSQL’s WAL is the mechanism that powers CDC.)

OLTP - Online Transaction Processing :

  • Handle High volume short, fast and day-to-day transactions
  • optimized for reads and writes per row
  • queries are fast and simple.
  • postgres is a good example
  • ACID gaurantee.