🔧 Do SLAs, Error Budgets, and Availability Metrics Include Maintenance Windows?

When it comes to service reliability, maintenance windows can be a gray area. Whether you're tracking uptime, setting SLOs, or managing customer expectations through SLAs, the question often comes up:

“Should scheduled maintenance count against our SLA? What about our error budget or availability metrics?”

Let’s unpack how scheduled (and unscheduled) maintenance affects your SLAs, error budgets, and availability calculations — and what best practices look like.

---

📜 SLA: Do Maintenance Windows Count?

Service Level Agreements (SLAs) are typically contractual commitments made to customers, promising a certain level of service availability (e.g., 99.9% uptime).

✅ Planned Maintenance Is Usually Excluded

Most SLAs exclude scheduled and communicated maintenance windows from downtime calculations. That means if:

• Maintenance was planned,

• Properly communicated in advance (often 24–72 hours), and

• Done within agreed-upon time windows (e.g., off-peak hours),

…it usually does not count against the SLA.

❌ Unscheduled or Overrun Maintenance May Count

However, if:

• The maintenance wasn't properly communicated,

• It ran longer than scheduled,

• It was done during peak usage without approval,

…it can count as downtime and lead to SLA violations or service credits.

---

🎯 Error Budgets: Are They Affected by Maintenance?

Error budgets represent the amount of failure or unreliability tolerated over a period, based on an SLO (Service Level Objective). If your SLO is 99.9% uptime per month, your error budget is about 43.2 minutes of allowed downtime.

🚫 Planned Maintenance Usually Doesn’t Burn Budget

If maintenance is pre-approved and doesn't disrupt users, it’s typically excluded from the error budget — especially in SRE frameworks that prioritize user-perceived reliability.

✅ User-Impacting Events Do Burn Budget

If users are affected — even during scheduled maintenance — some orgs choose to count it against the error budget. The key question is:

“Would a user notice or be blocked?”

If yes, it probably burns error budget. If no, it likely doesn't.

---

📈 Availability: Does Maintenance Affect It?

Availability is the actual measured uptime of your service over a specific period — typically expressed as a percentage like 99.95%.

Whether maintenance counts against it depends on how you define availability in your metrics.

🔸 User-Facing Availability

If your availability metric reflects user impact, planned maintenance that’s properly communicated is often excluded.

🔹 System-Level (Strict) Availability

If you're measuring raw service uptime (e.g., pings, monitoring checks), all downtime — including planned maintenance — might be included.

---

📌 Summary Table

Maintenance Type	Counts Toward SLA?	Burns Error Budget?	Affects Availability?
Planned & Communicated	❌ Usually Not	❌ Usually Not	❌* If defined that way
Unplanned or Overrun	✅ Yes	✅ Yes	✅ Yes
Poorly Communicated	✅ Yes	✅ Yes	✅ Yes

---

🧠 Best Practices

• 📑 Define everything explicitly: Make sure SLAs, SLOs, and availability metrics clearly state how maintenance is handled.

• 📣 Communicate proactively: Proper notification is key to excluding maintenance from SLAs and error budgets.

• 🎯 Focus on user impact: Base decisions on whether users are affected, not just whether systems are up or down.

• 🤝 Align across teams: Ensure engineering, product, and legal are aligned on how you track and report service health.