07Stage 7 of 8 · Operations
What does reality show?
Here you compare the design with the running system and gain insights for the next improvement.
You need
- Documented architecture
- Assessed risks
Who
Operations leads
with Platform team
How to proceed
- 1
Prepare rollout and recovery.
- 2
Check quality targets against the operating signals.
- 3
Examine unexpected behavior and return the findings to the team.
In the shop
Operations with goals and restart
- SLO
- Order confirmation visible within two seconds, with a monthly target.
- Timeout
- Time budget ends after one second; a late provider response is reconciled.
- Rollout
- New tax rule first for a small share of traffic, additive schema.
- Restore
- Restore in isolation and compare orders against the payment system.
This creates
- Operating signals
- Audited operational processes
- New findings
Tools for this:SLO & error budgetRecovery and restore exerciseAll under “Protect” →
Done when
- Rollout and recovery are prepared and rehearsed.
- Quality goals are measured against operational signals.
- Anomalies are fed back to the team.
Appropriate Building Blocks
Operations · Quality attributeObservabilityObservability combines signals that let you inspect a system's internal state.Operations · Decision spaceDeployment & rolloutDeployment and rollout distribute software changes to the runtime environment under control.Operations · ActivityRecoveryRecovery restores data and functionality after a failure using verified procedures.Operations · ApproachSLIs, SLOs & error budgetsSLIs, SLOs and error budgets connect measurable operational quality with targets and agreed deviations.Operations · MechanismTimeoutA timeout limits waiting time. Here you learn how the caller handles the resulting state.
Findings from here flow back into the draft. To: Design →