Communicate
Reports
Availability against your SLA target for any period: for every monitor or for one client's, emailed on a schedule, shared as a link, exported to CSV, or read from the API.
Overview
A report answers one question: did we keep our reliability promise for this period, and if not, why? It shows availability against your SLA target, the downtime that counted, the incidents behind it, and how each monitor did. Reports live in the Reports section of the dashboard.
Every workspace has a report called All monitors, which covers every monitor and is emailed weekly. On Growth and above you can add more reports, each covering the monitors you choose (for example one client's), with its own SLA target, schedule, recipients and branding.
How Uptime Is Calculated
Uptime is time-based, not the share of successful checks, so it doesn't change when you change a check interval and short outages aren't hidden at long intervals.
- Availability is the share of monitored time a service wasn't down, per monitor, shown to 3 decimals (99.950%).
- Downtime is the time an automatic incident was open: from the first failed check, confirmed from enough of the monitor's regions (a majority by default; see incident conditions), until it recovered. A failure in one region alone isn't downtime. Manual incidents don't count.
- Reports cover outages. A slow service, or one failing in only some regions, isn't down: degraded incidents stay out of reports, and one that turns into an outage counts from when the service went down. Deploy incidents and incidents on dependencies you follow stay out too. Manual incidents are listed, but never count as downtime.
- Planned maintenance isn't downtime, from the moment the maintenance window existed. A window created more than 15 minutes after it started is marked on the report, and the time before it was created still counts.
- Monitored time starts when a monitor was created, so a monitor added mid-month isn't credited with uptime for the days before it existed.
- The report's figure is the time-weighted average of its website and API monitors.
- Heartbeats (cron jobs and background tasks) are reported as check-ins received, not as availability.
Every report states the method it was calculated with under "How is this calculated?", so a reader doesn't have to take the number on trust.
Periods
| Period | Description |
|---|---|
| Weekly | Weeks start on the report's send day. Past weeks are the reports as they were emailed; this week shows the week so far. |
| Monthly | Calendar months, including the month so far |
| Quarterly | Calendar quarters, including the quarter so far |
| Range | The last 7, 30 or 90 days, or any range of days up to 400 |
Every period is compared with the period of the same length right before it. Days are cut in the workspace's timezone, shown on every report.
What's in a Report
| Section | Description |
|---|---|
| SLA verdict | Availability against the report's SLA target: met or missed |
| Summary | Availability with its change from the previous period, downtime against the downtime the target allows, heartbeat check-ins, incidents, time to acknowledge (MTTA) and time to resolve (MTTR) |
| Monitors | Each monitor's availability, downtime, incidents, p50 / p95 / p99 response times, checks and failed checks. Expand a monitor for its regions and a daily p95 chart per region. Monitors deleted since are marked. |
| Heartbeats | Check-ins received and missed per job |
| Incidents | Longest first, with the monitor, the cause (timeout, DNS, TLS, connection, 5xx, 4xx, failed assertion), the regions that failed, and a link to the incident. Downtime is also totalled by cause. |
| Planned maintenance | Maintenance windows on the report's monitors, with how long each ran in the period |
SLA Target and Slow Days
Each report is judged against an SLA target: 99%, 99.5%, 99.9%, 99.95% or 99.99%. The default is 99.9%; other targets are on Growth and above. Set it in the report's Schedule & delivery to the target in your contract or on your pricing page, so the verdict and the downtime budget mean something.
A report can also set a slow days limit: a p95 response time in milliseconds. Days above it are counted per monitor and region and drawn as a line on the latency chart. Slow days are reported only; they never affect availability or the SLA verdict.
Reports for Clients
Click New report to create a report that covers chosen monitors (website and API monitors and heartbeats, with a search and tag filter to pick them quickly) or monitors with tags. Agencies typically make one report per client, with the client's SLA target, a monthly schedule in the client's timezone, and the client's recipients.
A report of chosen monitors keeps that list: monitors added to the workspace later aren't included until you choose them. A report of monitors with tags covers every healthcheck and heartbeat with any of its tags each time it's built, so tagging a new service with the client's tag adds it to the next report. A report that's already been sent keeps the monitors it covered.
The Monitors filter narrows the report on screen to some of its monitors, for example the three public-facing ones a prospect asked about. It isn't offered on a report as it was sent. A share link created from that view keeps the selection.
Scheduled Emails
A report with a schedule is frozen when each period closes and emailed to its recipients: weekly on a weekday, or monthly or quarterly on a day of the following month, at a time in the timezone you choose. A frozen report never changes afterwards, so what a client received is what the report shows later. Weekly emails are on every plan; monthly and quarterly schedules and reports other than All monitors are on Growth and above.
- Each recipient gets their own email; recipients don't see each other. Add your own address to a client's report to get the same copy.
- A new workspace's All monitors report goes to the person who created the workspace. Manage everyone else in the report's recipients.
- Recent deliveries in Schedule & delivery shows what was sent to whom and when; a failed delivery can be sent again.
- The email shows the summary, monitors worst first, heartbeats, the longest incidents and planned maintenance, with a link to the full report.
Branding and White-Label
On Growth and above a report can carry the logo, name and accent color of the company it's for. The logo leads the email and the shared page, the name is used in the email's subject, and the accent color is used for buttons and links. Logos can be PNG, JPEG or GIF up to 2 MB (email clients don't all show SVG or WebP).
With white-label, Uptimy's logo and product footer are removed from report emails, and the shared page's browser tab shows only the company and the report. A small "Powered by Uptimy" stays.
Sharing
In the Share & export menu, Share link creates a public link to the report on screen and copies it. A past week, month or quarter can be shared, and so can a range of days, which keeps its dates. The week, month or quarter in progress can't. The shared page needs no sign-in and is served at:
https://app.upti.my/reports/{token}- A link keeps the monitors and the SLA target it had when it was created, so a monitor added later never appears on a link already sent.
- Links expire after 7, 30 or 90 days, or never, as you pick in the menu. Deleting a link under Share links revokes it. An expired or deleted link shows "Report not found".
- The shared page shows the full report, without links into your workspace. Its Print / Save as PDF prints a clean A4 document, in light colors even if the reader uses dark mode.
CSV Export
On Growth and above, Download CSV in the Share & export menu exports the report on screen as:
| Export | One row per |
|---|---|
| Summary per monitor | Monitor, after a line of totals |
| Per monitor per day | Monitor and day: availability, downtime, checks, latency percentiles, check-ins |
| Per incident | Incident: start, end, duration, cause, regions |
| Planned maintenance | Maintenance window and the time it ran in the period |
API, CLI and MCP
The same reports and report settings are available with an API key at https://reports.upti.my/v1/api/reports, under the same plan rules as the app. Logos and share links are set up in the app.
Reading them:
# The workspace's reports
curl -H "Authorization: Bearer upt_..." https://reports.upti.my/v1/api/reports/definitions
# September for every monitor, as JSON or CSV
curl -H "Authorization: Bearer upt_..." https://reports.upti.my/v1/api/reports/monthly/2026-09-01
curl -H "Authorization: Bearer upt_..." "https://reports.upti.my/v1/api/reports/monthly/2026-09-01?format=csv&rows=days"
# Any range of days for one report
curl -H "Authorization: Bearer upt_..." "https://reports.upti.my/v1/api/reports/range?from=2026-09-01&to=2026-09-15&definition=<report-uuid>"With uptimyctl (v1.6.0 or later):
uptimyctl reports list
uptimyctl reports get # last 30 days, every monitor
uptimyctl reports get --last 7d --report Acme
uptimyctl reports get --month 2026-09
uptimyctl reports get --quarter 2026-Q3 -o json
uptimyctl reports get --from 2026-09-01 --to 2026-09-15 --csv days > september.csvWith the MCP server, ask your assistant something like "Did we meet our SLA last month?"; it uses list_reports and get_uptime_report. It can set reports up too: "Add a monthly report for Globex's monitors at 99.5% and email it to cto@globex.com" uses add_report, and update_report changes one. With uptimyctl, reports are part of the workspace file.
Plans and Permissions
| Feature | Free and Starter | Growth and above |
|---|---|---|
| All monitors report, weekly email, past weeks | Yes | Yes |
| This week, months, quarters and ranges | No | Yes |
| Reports for chosen monitors, monthly and quarterly emails | No | Yes |
| SLA targets other than 99.9%, slow days | No | Yes |
| Share links, CSV, the Monitors filter | No | Yes |
| Branding and white-label | No | Yes |
Every workspace member can read reports. Changing a report's schedule, recipients, target or branding, creating reports and creating share links need the permission to update workspace settings (admins).
Match the SLA You Promise
Set each report to the SLA target in the contract it reports on. The verdict and the allowed downtime are only useful when the target is real.