New: Cookbooks and AI ExplanationsStep-by-Step recipes to solve problems connected to Roadmaps and Cheat Sheets. Need more details? Use AI buttons for structured and simple explanations with concrete examples throughout the whole platform.Take a look
Every update to your lakehouse table rewrites whole files, and the write job is starting to crawl.
What you'll have at the end
The table's write mode set to match its update pattern, with write and read times measured before and after
You need
An Iceberg table, format version 2 or higher, already taking individual row deletes, updates, or merges from a live pipeline, plus a way to replay a representative batch of those same writes on demand.
Not covered
Choosing a catalog, standing up the pipeline that feeds the table, or the newer deletion-vector mode that folds both strategies into one.
Set a snapshot expiration policy before old versions eat your storage bill
once compaction is rewriting files on a schedule, the snapshots holding the pre-compaction versions need their own expiration policy or storage keeps growing anyway
A lakehouse query keeps slowing down as manifest bloat builds up
if reads stay slow even after compacting the delete files, the next place to look is the manifests themselves
Checked 26 Aug 2026
Part of the Data Lakehouse cookbook