Migrations & Catalogs — Week 3 CheckWeek 3 of Lakehouse & Table Formats: migration mechanics and the catalog wars — from Kaltura's Redshift escape to LinkedIn's OpenHouse.

Kaltura was running scheduled VACUUMs and deleting data to keep Redshift alive. What does that pattern signal?

Kaltura bridged the transition with Delta tables exposed to Redshift via symlink manifests. Why does that bridge matter in a migration?

Peloton's CDC lakehouse hit AWS Glue's one-million-resource ceiling. The general lesson:

LinkedIn's OpenHouse cut table onboarding from 2-3 weeks to seconds. The architectural move that made it possible:

With Iceberg consolidating as the format standard, where does the Späti 'ICE stack' analysis say lock-in re-enters?

Before committing to a managed Iceberg offering, the one compatibility question to verify: