MySQL and PostgreSQL are the goto SQL databases for a huge number of services. But what if your service generates billions or even trillions of records? This talk helps listeners decide when applications work better with a SQL data warehouse.
Using MySQL and ClickHouse as examples, we'll work through use cases like transaction processing and computation of aggregates to see where each shines. Along the way we'll explore key technical differences:
- row vs. column storage
- query parallelization
- sharding and replication
- concurrency support
- transaction models
Finally, we'll discuss how to handle use cases that require the capabilities of both. Listeners will leave with clear criteria and a deeper understanding of database internals that enable them to make the right choice(s) for their own use cases.