HTAP
Hybrid Transactional/Analytical Processing — a database architecture that runs both transactions (OLTP) and analytics (OLAP) on the same system, without separate ETL to a data warehouse. Goal: analytics on fresh data in real time.
Term coined by Gartner ~2014. The problem it solves: traditionally OLTP (fast small writes) and OLAP (heavy aggregations) run on separate systems → analytics lag hours behind. HTAP systems keep both a row-oriented (OLTP) and a column-oriented (OLAP) representation, often synced internally. Examples: TiDB (TiFlash column store), SingleStore, Google AlloyDB, SAP HANA. Trade-off: more complex, harder to optimize both workloads at once. Modern trend: separation of storage/compute makes HTAP more practical. Competes with the "ZeroETL" concept from cloud providers.