Skip to content
LakeBench
ProblemsCommunityPricing
Sign inStart practicing
Back
  1. Home
  2. Interview prep
  3. Observability vs monitoring

Data quality · Quality Foundations

Observability vs monitoring

Mediumquality-09
observabilitymonitoringfreshnessvolumeschema

Question

What is the difference between data monitoring and data observability?

Solution

Monitoring watches known signals against thresholds you already defined: "alert if job fails," "alert if freshness > 2 hours," "alert if row count < 1,000."

Observability is the broader ability to understand system health from telemetry so you can debug unknown failures, not only the alerts you predicted.

monitoring:     is fct_orders fresher than 2h?  yes/no
observability:  freshness + volume + schema + null rates
                + distribution shifts + lineage blast radius
                + recent deploy / DAG run context

Data observability pillars (common interview list) 1. Freshness 2. Volume 3. Schema 4. Distribution (null rates, uniqueness, value shapes) 5. Lineage / downstream impact

Practical difference

A monitor fires: "row count dropped 80%." Observability helps you see *why*: upstream source empty, schema rename broke the join, or a filter bug in yesterday's deploy, and which dashboards are affected.

Interview tip: "Monitoring = predefined alerts; observability = enough telemetry and context to diagnose unknowns." List the pillars.

PreviousNext