Maintenance Window Management Tools
A database upgrade or a CDN migration will trip your monitors — as it should, if it’s a surprise. Schedule a Maintenance Window and AmICited classifies checks taken during it as maintenance: the monitor’s state and consecutive-failure count freeze in place, and no incident opens, updates or resolves off them. This covers active HTTP, TCP, DNS and Transaction monitors; passive heartbeats aren’t matched by a window yet, so pause those directly.
Brands tracking how AI answers name them — sampled daily across every surface their buyers ask on.
Read the case studies →Maintenance window scope
A window can cover every active monitor in the workspace, every monitor on one domain, or just the single monitor affected by the work — a checkout flow during a payment-provider migration, say — with a start and end time or a repeating schedule for routine maintenance, evaluated in UTC.
- Workspace, domain or monitor scope — suppress everything, everything on one domain, or just the specific monitor the work touches; a monitor-level scope requires a domain scope first.
- One-time, daily or weekly recurrence — a single migration window, or a schedule that repeats the same duration every 24 hours or seven days, capped at seven days per occurrence.
- Incidents stay quiet, state is preserved — no false alert fires, and the monitor’s prior state carries through rather than flipping to a false healthy or failed reading.
- SLA math excludes it automatically — planned downtime never counts against your uptime target or shows up in a customer-facing SLA report.
Maintenance monitoring classification
A maintenance window doesn’t stop checking. Probes still fire, and the raw result is stored, but the outcome is classified as maintenance instead of up or down — so you keep evidence from the work window without it contaminating reliability numbers. When the window ends, normal state transitions resume on the next check; extend the end time if the work overruns, or a real failure can start counting once the schedule no longer covers it.
- Checks still run and are stored — probes during a window are recorded, just classified as maintenance rather than up or down.
- Excluded from every rollup — uptime percentages, response-time averages, availability bars and monthly SLA math all drop maintenance checks.
- An “In maintenance” badge — shown on the monitor list and detail header the moment a matching window is active.
- Extend before it lapses — if work overruns the scheduled end, extend it before the next check, or that check is judged normally.
Maintenance window SLA reporting
A maintenance window scheduled ahead of time keeps your SLA Report honest — customers and stakeholders see uptime against your target with planned work already excluded, not padded or disputed after the fact.
- Maintenance excluded by default — no manual adjustment needed when compiling a monthly report.
- Exportable evidence — download CSV for any month, ready for customer or internal reporting.
- Target vs. actual, monthly — see whether you met your SLA target with the numbers that should count.
false incidents from work you already scheduled
A window enabled ahead of time means the exact downtime you planned never gets treated as an outage by monitors, incidents or SLA math.
See Uptime Monitors