🛡️Data Observability
Also known as Data Monitoring, Data Pipeline Monitoring
Watching data all the time, and calling out when it looks odd.
Think of it like
It is like a smoke alarm for your data. You do not say where the fire will be. It just notices smoke.
Example
On Thursday, an orders table gets 40% fewer rows than usual. No check covered it, but the monitor flags the drop within the hour.
Why it matters
Checks only catch the problems you predicted. Observability catches the ones you did not.
Definition
Common mistake
Key takeaways
- 1Learns normal. Builds a baseline for each table.
- 2Watches signals. Freshness, volume, schema and values.
- 3Alerts on change. Flags what no one predicted.