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
The job died at row eight thousand, and rerunning it just doubled everything that already loaded.
What you'll have at the end
A rerun that leaves the table with the same rows and totals as the first run
You need
A target orders table already holding rows from past runs, including a first_seen_at column that should never change once set, plus the ability to pull the same source batch again for testing.
Not covered
Deciding whether tonight's pull grabs the whole source table or only what changed since last time.
Choose full reload or incremental for a new source
go there first if you haven't decided whether tonight's pull grabs the whole source table or only what changed.
Swap slow row-by-row inserts for a bulk COPY load
go there if the staged batch itself runs into the millions of rows and even one set-based INSERT is too slow.
Checked 25 Aug 2026
Part of the ETL (Extract, Transform, Load) cookbook