A fact row arrives before its dimension row. How is late-arriving data usually handled?
Why A is the answer
Insert a placeholder dimension row and update it later
A placeholder keeps the fact usable until the real dimension arrives.
In detail
- A placeholder keeps the fact row valid.
- The dimension row is updated once it arrives.
- Dropping the fact would understate the totals.
Use case
An illustrative scenario, not a real incident.
A sale arrives for a product not yet in the catalog. A placeholder product keeps the sale, and the real details replace it later.
See it in different ways
stateDiagram-v2 [*] --> Bronze: raw data as it arrived Bronze --> Silver: cleaned and validated Silver --> Gold: modeled for business use Gold --> [*]
stateDiagram-v2 state "Fact arrives" as s0 state "Placeholder row" as s1 state "Real dimension row" as s2 [*] --> s0 s0 --> s1: dimension missing s1 --> s2: details arrive s2 --> [*]
Key point: Placeholders keep late facts usable.
Why the other options are wrong
- B. Stop the pipeline until someone fixes it by hand
- Stopping the pipeline delays everything.
- C. Keep it in a file outside the warehouse
- A file outside the warehouse leaves it out of reports.
- D. Delete the fact row and wait for a resend
- Deleting loses a real fact.