Data Warehouse
Learned in CS451.
Applications have two very different access patterns against data:
- OLTP — user-facing, low-latency, concurrent, small reads/writes
- OLAP — back-end, batch, full-table scans, complex analytics
Two patterns, one database. To satisfy both needs, we use a data warehouse — a separate analytics store populated from the OLTP system, so that heavy OLAP queries don’t interfere with user-facing traffic.
