Website availability monitoring and SLA reports

Uptimia counts an outage only after other locations confirm it. Every percentage you send comes with the incidents and check log behind it.

Start free trial → See what gets counted
30-day free trial No credit card 171+ checkpoints

Which minutes reach the percentage

A month of checks includes one-location blips, planned deploys and real outages. Only the confirmed outages belong in the percentage you send.

Only the two confirmed incidents reach the number. If a client asks where 99.93% came from, you can show them both incidents with their durations and the deploy listed separately as maintenance. The Frankfurt blip is kept in the check log and adds no downtime.

How the percentage is calculated

These rules decide which minutes count toward the percentage. The record your client reads shows where each one applied.

  • Confirmation A failed check is re-run from up to 3 other locations before an incident opens, so a bad route at one probe stays out of your client's report.

  • Maintenance Planned minutes leave the measured time, counted as neither downtime nor uptime. Schedule the deploy inside a window and it uses none of the SLA budget.

  • Period edges An outage that started the night before counts only its minutes inside the period, and a month with any incident shows 99.99% at most.

  • Paused monitors A paused monitor drops out of the fleet average. It is never counted as up, so switching one off can't raise the number.

A month with an incident is never rounded up to 100.00%. How false alarms are filtered out → 99.99%

Where your availability appears

The dashboard and the scheduled report email show the same record, so people who never log in see the numbers too.

Scheduled report

The report your client receives

The email opens with a verdict and lists every monitor underneath. A client can read it top to bottom without a dashboard, and file the attached PDF wherever the SLA is kept.

  • Summary at the top. A one-line verdict on the period, then the totals for monitors and for critical and non-critical incidents.
  • One row per monitor. Uptime, average response, downtime and incident count side by side. A monitor that had an incident in the period is marked red, and the rest are green.
  • Incidents under each monitor. Each is listed with the error and how long it ran, and maintenance windows are listed apart, so planned work never reads as an outage.
Explore scheduled reports →

Availability chart and percentage

The percentage and its chart for any range you pick, with every incident in that range listed beside it.

Cumulative downtime and recovery time

Total downtime and mean time to recovery for the incidents in view, a CSV export, and a PDF, HTML or public link for any single incident.

Public uptime with a 90-day bar

Each monitor's state, uptime percentage and a 90-day history bar on a public URL or your own domain, where customers can check the current status themselves.

Explore status pages →

Send clients an availability reportbuilt from every confirmed outage.

Each check runs from one of 171+ locations, and every confirmed minute of downtime is filed.

Start your free 30-day trial →
30 days free no credit card cancel anytime

When the first alert goes out

Every confirmed minute goes on the record, so your number depends on how soon someone starts the fix. Check frequency, outage confirmation and alert delay decide when the first alert goes out.

Checks as often as every 30 seconds

The check interval sets how soon an outage is seen and how precisely it is measured. On a 5-minute interval you cannot prove a 99.99% month, because a single missed cycle is longer than the whole allowance.

Up to 3 locations must agree

Before an incident opens, the failed check is re-run from other locations. With the setting at 3, network trouble at one probe cannot page anyone or count as downtime. Lower it to 1 and a single backup check from another location is enough, which alerts you sooner but lets more false alarms through.

Alert delay leaves the downtime unchanged

The delay only holds back the notification. The incident opens the moment the outage is confirmed, so the downtime counts the same either way. An outage that ends inside the delay is still recorded, and nobody is woken for it.

  • Alert channels. Email, SMS, Slack and the other channels your team already uses.
  • Incident details. Each incident lists which locations confirmed it, when, and what they saw.
  • The recovery notice includes the duration. Everyone who was paged learns the site is back up and how long the outage ran.
Explore downtime alerts → Explore website uptime monitoring →
Recovered after 6 min · all-clear to everyone alerted 6 min on record
Locations

Checks from 171+ locations

An error seen from one location may come from that location's route to your site. Uptimia re-checks from other locations before a minute of downtime is written, and stores every check with its checkpoint location.

  • Every check is logged. The log keeps the time, location and result of each one, so "6 minutes down" can be read check by check.
  • Failures need a second location. A failure that other locations cannot reproduce is logged but not counted in the percentage.
  • Export the raw data. CSV from the logs, PDF or HTML for one check, and the REST API for your own reports.
Explore multi-location monitoring →
171+ checkpoints · 70+ countries check log

Also included

Scheduled reports

Daily, weekly, monthly, quarterly or yearly to the people you name, or a one-off report for any exact date range. New accounts start with summaries already going to their own address.

White-label reports

Your logo and colours on the report before it reaches the client, with the Uptimia footer removed.

Maintenance windows

Alerts stay quiet during planned work, and the window is kept in the record, listed apart from real incidents.

Read-only client seats

Invite a client as a read-only member and limit the seat to the monitor groups that hold their sites.

Bulk create from a pasted list

Paste one URL per line, preview every row, and create uptime, SSL, domain, virus or speed monitors in one batch.

Bulk actions

Select any set of monitors and pause, resume, delete or schedule maintenance for all of them in one step.

How teams use the availability record

Each example shows the settings, what Uptimia recorded and what happened next.

Example 1Agency

A monthly report for a client

The contract says 99.9% a month, and the client wants it in writing without logging into anything. Six monitors cover their sites. In March the shop was down once, for 18 minutes.

Report settings FrequencyMonthly · 1st Monitors6 of 34 Email contentAll incidents · PDF ThemeAgency logo + colours
The record and what happened →
In the record
One row per monitor with uptime, downtime and incident count, and the shop's 18 minutes listed underneath with the error and how long it ran.
What happens
On the 1st the email goes out with your logo and colours and the PDF attached. If the client wants the live number in between, a read-only seat limited to their monitor group shows their own sites and nothing else.
Example 2SaaS

Checking an SLA budget mid-month

Your SLA allows 21 minutes 36 seconds of downtime in a 30-day month. So far Uptimia has recorded two confirmed incidents, 9 minutes and 4 minutes. The weekly deploy runs inside a maintenance window, so none of it is downtime.

Monitor settings Check frequencyEvery minute Outage confirmation3 locations MaintenanceWeekly · deploy
The record and what happened →
In the record
The availability chart shows the month-to-date figure. Incident history lists both incidents with their durations, plus cumulative downtime for the incidents in view. Subtract that from your allowance, or let the free SLA calculator do it. The month's incidents export as CSV for the quarterly review.
What happens
The team is paged for each incident once it is confirmed, and the recovery notice says how long it ran. Nobody is paged for the deploy, because alerts are silenced inside the window.
Example 3Online shop

Checking a customer's downtime claim

A customer says the shop was unreachable for most of Tuesday afternoon, while your team remembers a 6-minute blip. You don't need to set anything up for this, because the record was written while the outage was happening.

Monitor settings Check frequencyEvery minute Outage confirmation3 locations Date rangeTuesday afternoon
The record and what happened →
In the record
The timeline shows where the 6 minutes sat. The incident lists the locations that confirmed it and what they saw, and the check log shows every other location getting an answer for the rest of the afternoon.
What happens
Export the incident as a PDF or send the customer a public link to it, so they can check the times and locations for themselves.

Check the numbers yourselfon your own sites.

Add them today and pick who gets the report and how often.

Start your free 30-day trial →
30 days free no credit card cancel anytime

How website availability monitoring works

Uptimia checks your website from 171+ locations in 70+ countries and confirms each failure from other locations before it counts. Availability is the share of monitored time outside confirmed incidents, shown per monitor as a percentage and a chart, with every check logged behind it and scheduled reports for anyone who needs the proof.

01What is website availability?+
Availability is the share of time your website answered correctly over a period, expressed as a percentage. 99.9% over 30 days means 43 minutes 12 seconds unreachable. It is measured from outside your infrastructure, because a site that is up for your servers can still be unreachable for your visitors.
02How does Uptimia calculate the percentage?+
Every check is stored with its location. When enough other locations confirm a failure, an incident opens; when checks pass again, it closes. Availability is the monitored time outside confirmed incidents divided by the monitored time, with maintenance windows removed from both sides of the division. An incident that spans the edge of the period contributes only the minutes inside it, and a period with any incident at all shows 99.99% at most, never a rounded-up 100.00%.
03What is the difference between uptime and availability?+
In practice the words are used interchangeably. When people separate them, uptime monitoring means the live question, "are we down right now?", answered with an alert as soon as the site fails. Availability means the measured record of how often you were up over a period. The same Uptimia monitor sends the alert and keeps the record. What website availability is · How it differs from uptime
04What availability target should I aim for?+
Pick a target based on what an hour of downtime costs your business. In a 30-day month, 99.9% allows 43 minutes 12 seconds of downtime, 99.95% allows 21 minutes 36 seconds and 99.99% allows 4 minutes 19 seconds. Each extra nine costs more engineering than the last, and it needs a tighter check interval too, because a 5-minute interval cannot prove a 99.99% month when one missed cycle is longer than the whole allowance. Uptimia's free SLA and uptime calculators turn any target into allowed minutes. SLA calculator · Uptime calculator · Downtime cost calculator · Allowed downtime for every target
05Do maintenance windows affect my availability?+
A window's minutes are taken out of the time being measured, so planned work is counted as neither downtime nor uptime. Any incident open when the window starts is closed, alerts are silenced for its duration, and checks resume when it ends. When a report is set to include maintenance, windows are listed apart from incidents, so a client can see which minutes were planned. Windows can be one-off or recur daily, weekly or monthly.
06Does the alert delay change the recorded downtime?+
No. An incident opens the moment the outage is confirmed, and that is when the downtime starts counting. The alert delay only holds the notification back, so a monitor set to wait a few minutes before paging anyone records exactly the same downtime as one that pages at once.
07What counts as "down"?+
Whatever you tell the monitor to treat as a failure. By default an HTTP monitor counts 4xx and 5xx responses, connection failures, timeouts, redirect errors and certificate errors as down; you can switch individual response classes on or off per monitor, set the timeout, and add keyword rules so a page that returns 200 with an error message in the body still counts as down.
08Can a client check the live number without asking me?+
Yes. Invite them as a read-only member and limit that seat to the monitor groups that hold their sites. They see those monitors' availability, incidents and check logs and nothing else, and they cannot change a setting.
09Which monitors appear in scheduled reports?+
Uptime, SSL, domain, speed, RUM, virus and transaction monitors, plus blacklist monitors when a report covers all monitors. Heartbeat, server, DNS and API monitors keep their own incident history and check logs in the dashboard but are not part of the scheduled report.
10How long does a public report link stay live?+
A public report link is a snapshot and is removed after 30 days. Export the PDF for anything you need to keep longer, or let the scheduled email deliver a fresh copy each period.
11Can I pull the raw data myself?+
Yes. The REST API returns each monitor's incidents and check logs, incidents can be exported as CSV from the history view, and every single check or incident can be exported as PDF or HTML from the dashboard.
12What does the free plan include?+
The free plan monitors 1 website at 5-minute intervals, with alerting, scheduled reports and 7 days of history. Paid plans keep 90 days to 2 years of history, and Professional and above check as often as every 30 seconds and can white-label report emails. The 30-day trial opens every monitor type, with no credit card.

Availability reports backed by check logs

Outages are confirmed and filed as they happen, and the monthly report goes out on schedule.

Start free trial →
30-day free trial Every check logged Reports included No credit card
Your first availability report can go out this month.