Skip to content
LakeBench
ProblemsCommunityPricing
Sign inStart practicing
Back
  1. Home
  2. Interview prep
  3. SLA vs SLO vs SLI

Data quality · Extra High-Value

SLA vs SLO vs SLI

Mediumquality-19
SLASLOSLIfreshnessreliability

Question

What is the difference between SLA, SLO, and SLI for data products?

Solution

These three terms come from reliability engineering and map cleanly to data platforms:

SLI (Service Level Indicator): the raw metric you measure. Examples: freshness lag in minutes, successful publish %, test pass rate, availability of a table endpoint.

SLO (Service Level Objective): the internal target for that SLI. Example: "99% of weekdays, fct_orders freshness lag <= 60 minutes."

SLA (Service Level Agreement): the external/business promise, often with consequences if broken (credits, escalation, contractual terms). Example: "Customer analytics feed delivered by 09:00 with max 2h lag."

SLI:  lag_minutes = now - max(loaded_at)
SLO:  lag_minutes <= 60 for 99% of weekdays
SLA:  contractual promise to stakeholders about that delivery

How teams use them

You instrument SLIs, manage to SLOs, and only put SLAs on what you can actually meet. An SLA without SLIs is marketing fiction.

Interview tip: "SLI = measure, SLO = target, SLA = external promise." Give one freshness example that chains all three.

PreviousNext