What's in a report

Every monthly SLA report follows the same structure, so a client who has seen one can read the next in a minute. This article walks through each section and, more importantly, what each number actually claims.

The summary block

The first thing the client sees is a block of six numbers describing the whole month:

• Uptime vs the SLA target — the measured uptime percentage next to the target set on the client's record, so whether the SLA was met is visible in the first second.
• Checks — how many probes ran against the client's systems during the month. This is the size of the evidence behind everything else.
• Incidents — how many incidents opened during the month.
• p95 response time — the response time that 95% of checks came in under. It's a fairer picture than an average, which one slow hour can distort; see Percentiles and p95 explained.
• Downtime minutes — the month's unavailability expressed as minutes, which lands harder than a percentage; see What is downtime?.
• Measurement coverage % — how much of the month was actually being measured. This one deserves its own section below.

The per-system table

The summary averages across everything, so the report then breaks the month down system by system: one row per monitored system with its own numbers. This is where a client sees that the 43 downtime minutes all belong to one backup server, not to their public website — and where you see which system to talk about in the review meeting.

The incident log

A chronological log of the month's incidents: what happened and when. It matches the alerts your team received at the time, which is exactly the point — the client is reading a record, not a reconstruction.

Coverage disclosures

Measurement coverage is the percentage of the month during which the client's systems were actually being checked. 100% means every monitor existed, was enabled, and reported for the entire month. Coverage drops when a monitor was created mid-month, paused for a stretch, or stopped reporting.

The report states this number because an uptime figure is only as strong as the measurement behind it: 100% uptime at 40% coverage means "no failures while we were looking", which is a much weaker claim. Disclosing coverage keeps the report honest — and defensible when a client asks hard questions. The report archive shows the same coverage figure before you download, so you're never surprised by it.

The small print

Clock times in the report are printed in your workspace timezone, and a footnote at the bottom says so explicitly — no ambiguity about what "03:12" means. Day boundaries follow each monitor's own timezone; Report language and timezone explains the difference. Finally, every report carries a small "Powered by UpAll" footer, the same as portals and status pages.