Comparisons

Pairs that get conflated in real conversations. Neither column wins — what decides is the requirement, and each record leads with the confusion, because the confusion is the reason the record exists at all.

OLTP vs OLAP Batch vs Streaming ETL vs ELT Parquet vs Avro Lake vs Warehouse vs Lakehouse CDC vs Polling Event modelling vs Snapshot modelling Full refresh vs Incremental Lambda vs Kappa Order grain vs Order-line grain

Lake vs Warehouse vs Lakehouse

What people get wrong about this pair

The lakehouse is sold as making the choice obsolete. It narrows the gap by adding a metadata and transaction layer over files — which is a real advance — but the trade between openness and integration has not disappeared.

Data lake

Use it when

Cheap durable storage for anything, in any shape, that you may want to reprocess later.

Data warehouse

Use it when

Fast, governed, concurrent SQL over structured data with real transactions.

Dimension	Data lake	Data warehouse
Data shapes	Anything, including what you cannot parse yet	Structured, schema decided at write
Transactions	None natively — files are just files	Yes
Governance	Whatever you build on top	Built in, at table and often column level
Cost of storing	Lowest available	Higher, and coupled to the product
Query performance	Depends entirely on layout	Depends on layout, but the system helps
Openness	Files you can read with anything	Varies; the data may not be readable outside the product
The lakehouse position	Adds a table-metadata layer giving files snapshots, atomic commits and schema evolution	Increasingly reads and writes open table formats in place