Monitor configuration set per endpoint

Each Uptimia monitor keeps its own request, failure rules, check timing, recipients and locations. In one 12-minute restart, the checkout texts the shop owner within the first minute and staging sends nothing.

Start free trial → Compare alerts during a restart
30-day free trial No credit card GDPR-ready

Alerts per monitor during a restart

A 12-minute restart takes down a brochure site, a checkout and a staging box at once. Their monitors use different check timing and recipients. The shop owner gets a text about the checkout and one e‑mail about the brochure site. Staging sends nothing.

The checkout is checked every 30 seconds with no alert delay. The shop owner gets a text inside the first minute and an all-clear as soon as the page is back. The brochure site runs every 5 minutes on the default delay, so one e‑mail goes out once the failure has lasted longer than the delay, and nobody gets a text. Staging holds its alert for 15 minutes, and this restart was over before then, so nothing was sent.

What each editor setting controls

The screenshot shows the Checkout page monitor open in the editor. Each numbered marker matches a note beside it.

  1. 1
    The request

    URL, method, headers, basic auth and a body. With these the check gets past an endpoint that refuses an anonymous GET. Web, e‑mail and network sub-types cover HTTP, SMTP, POP3, IMAP, TCP, ping, DNS and UDP.

    Monitoring behind a login →
  2. 2
    What counts as down

    Keyword rules that must or must not match, the error classes that count, redirects and the timeout. A 200 response carrying an error page fails the check.

    Explore website error monitoring →
  3. 3
    Check frequency and alert delay

    Check frequency from 30 seconds to 24 hours, an alert delay for servers that recover on their own, and the recovery notice.

    Explore check intervals →
  4. 4
    Who is paged

    Contacts, teams and members picked for this monitor only. Each is reached on their own channel and inside their own alerting hours.

    Explore downtime alerts →
  5. 5
    Outage confirmation

    1 to 3 other checkpoints re-run a failed check before an incident opens, so a failure caused by a bad network route at one location does not alert anyone.

    Explore false-positive prevention →
  6. 6
    Where the checks run

    Every checkpoint, a built-in profile such as Europe, or a hand‑picked set of 6 or more.

    Explore multi-location monitoring →
  7. 7
    How it is organised

    Groups that filter the dashboard and limit teammates' access, a note shown on the incident, templates for the next monitor, and maintenance windows for planned work.

    Explore multi-site monitoring →

Where your settings show up

What you set in the editor shows on the incident and is available through the API.

Changed cards marked in the editor

The Advanced tab shows a green dot while any of its settings differs from the default. Each changed card has an edited chip with a one-click reset, so a teammate opening the monitor sees what was changed.

Your note on the incident page

Write up to 500 characters, such as who to call first or a runbook link. The note appears on the incident page, in alert e‑mails and in webhook payloads.

Monitor settings over the API

Method, body, headers, timeout, redirects, keywords, error classes, interval, locations, contacts and groups, with an API key from your account. Outage confirmation and basic auth are set in the editor only. An optional unique ID lets you look a monitor up by a reference your own system already holds.

Set the request and failure rules per monitor.Then choose the timing and who gets alerted when it fails.

Checkpoints in 70+ countries run the monitor with the settings you chose.

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

When an outage triggers an alert

A failed website check must be confirmed by other checkpoints and then outlast the alert delay before anyone is alerted. Both are set on the monitor. Recipients are picked per monitor, and each person keeps their own alerting hours.

Confirmed by other checkpoints first

1, 2 or 3 more checkpoints re-run the failed check and must also find it down before an incident opens. Keep 3 for a public site. Use 1 or 2 when a monitor runs from a short hand‑picked set and you want the alert sooner.

An alert delay after confirmation

Hold the alert for up to 30 minutes after the outage is confirmed. Leave it at zero for anything customers use. A server that restarts itself gets a longer delay, and you hear about it only if it stays down.

Recipients with their own alerting hours

Pick contacts, teams and members per monitor and reach them by e‑mail, SMS, chat, incident tools or webhooks. Each person sets alerting hours in their own timezone. An alert outside those hours is held until the window opens.

  • Recovery notice. Everyone who was alerted gets the all-clear with the outage duration. You can switch it off per monitor.
  • Note on the incident. Up to 500 characters, shown on the incident page, in alert e‑mails and in webhook payloads.
  • Maintenance windows. Checks pause and alerts stay muted for the length of the window. Run it once or repeat it daily, weekly or monthly.
Explore false-positive prevention → Explore check intervals →
Recovered · all-clear to everyone alerted duration

Settings by monitor type

Name, recipients, groups, notes and maintenance windows are on every type. Beyond those, each type shows only the controls that apply to it, plus a few of its own.

171+ checkpoints · 70+ countries EU

Choose where each check runs

Check a public site from all checkpoints, one location per check. Keep a regional shop on checkpoints in its own region. You pick each monitor's checkpoints from 171+ in 70+ countries.

  • Built-in profiles. Europe, North America, Asia, Middle East & Africa, South America, Oceania and EU / GDPR, and you can save your own.
  • EU accounts start on Europe. A new monitor on an account registered in the EU defaults to European checkpoints.
  • Hand-picked sets. Choose 6 or more. The checkpoint addresses are published as one IP per line or as an nginx geo block. To allowlist Uptimia, admit the full list.
Explore multi-location monitoring →

Also included

Monitor templates

Save any monitor as a template. The next monitor needs only a name and a URL, and every other setting carries over except stored passwords, which you enter again.

Checkout page · templatekeyword ruleevery 30 sSMSEurope
→
name + URL
Template → name + URL

Create monitors from a pasted list

Paste one line per site, with an optional name after a comma. Every row gets the same settings. Works for uptime, SSL, domain, virus and speed monitors.

https://caldmont.comhttps://caldmont.com/pricing,Pricinghttps://shop.caldmont.com,Shop… 39 more url,name on each line

Maintenance windows

Set a window for one monitor, a group or a whole type. Checks pause and alerts stay muted while you deploy.

MonTueWedThuFriSatSun
Sunday 02:00–03:00 Checks pausedalerts muted

Clone a monitor

Copy an uptime monitor with its settings and point the copy at the next URL, for example a second store or the same shop in another region.

Checkout page
Checkout page (copy)shop.caldmont.de/checkout
Same settings, new URL

Monitor groups

Group monitors by client or environment. Filter the dashboard by group, aim a maintenance window at one, or limit a teammate to the groups they look after.

ProductionStagingClient A
Dana · Editor · 2 groups Dashboard filteraccess scope

Multi-step API monitors

Chain up to 15 requests, pass extracted values between steps, and assert on status, headers, JSON and timing.

1POST /auth → {{token}}2GET /orders3status 200 · under 800 ms

Example monitor configurations

Setups that come up often in support. Each lists the exact values and what happens when the endpoint breaks.

Example 1Uptime monitor · HTTPS

Database error page served with 200 OK

The shop's app catches its own database errors and shows a "We'll be back shortly" page with a 200 status. If you check only the status code, that looks healthy. This monitor reads the body, finds the footer string missing and the error text present, and fails the check.

Monitor settings MethodGETMust exist"© Example Ltd"Must not exist"We'll be back shortly"FrequencyEvery 1 minuteAlert delay0
What the check sees and what happens →
What the check sees
The check log records a keyword error. Whoever opens it knows the server answered and the page content was wrong.
What happens
Three more checkpoints confirm the failure, and with no alert delay the alert goes out at once. When the footer string is back, the incident closes and the recovery notice reports how long it lasted. Explore keyword monitoring →
Example 2Uptime monitor · HTTPS

Token-protected API health check

The endpoint answers 401 unless the request carries a token. The monitor sends a long-lived token in a header on every check. When the database behind the API times out, the endpoint still answers 200 and puts "status":"degraded" in the body.

Monitor settings HeaderAuthorization: Bearer ••••HeaderAccept: application/jsonMust exist"status":"ok"Timeout10 sFrequencyEvery 30 seconds
What the check sees and what happens →
What the check sees
A check on the status code alone would pass this 200 response. The keyword rule catches the missing "status":"ok". A 401 also fails the check, so a revoked token opens an incident too.
What happens
The alert goes out once the failure is confirmed. For an API that wants a sign-in first, an API monitor posts the credentials in step 1 and reuses the returned token in later steps. Explore API monitoring →
Example 3Uptime monitor · HTTPS

Internal service behind an IP allowlist

A VPN gateway drops traffic from addresses it doesn't know. Checks from those addresses time out, and the monitor reports outages that users never had. Add the full published checkpoint list to the allowlist, IPv4 and IPv6.

Monitor settings Targethttps://gw.example.com:8443/healthzLocationsHand-picked · 6 checkpointsTimeout10 sConfirmation2 backup checksRecipients#infra
What the check sees and what happens →
What the check sees
Checks from any checkpoint now get through the firewall, so a timeout means the service stopped answering. With only 6 checkpoints in the set, 2 backup checks give enough confirmation.
What happens
Uptimia builds both IP files from the checkpoint list on every fetch. Pull one into the firewall config on a schedule and reload, and a new location reaches the allowlist without a manual edit. The checkpoint IP list →
Example 4Uptime monitor · HTTPS

Staging redeploys without phone alerts

Staging rebuilds several times a day, and each restart takes it offline for up to 8 minutes. Most redeploys come back on their own, and for the occasional one that doesn't, a message in the team channel is enough.

Monitor settings Basic authstaging · ••••Alert delay15 minutesRecipients#staging onlyRecovery noticeOffMaintenanceDaily 02:00–03:00
What the check sees and what happens →
What the check sees
Each restart still shows in the check log as failed checks, so staging's history stays complete.
What happens
A restart that recovers within the 15 minutes sends nothing. On the default 5-minute delay, the same restart could have alerted someone. A deploy that stays broken alerts #staging, and the nightly rebuild runs inside its maintenance window with checks paused. Explore maintenance windows →

Copy one of these setupsto your own monitors.

Save a monitor as a template, and the next one needs only a name and a URL.

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

How monitor configuration works

Every Uptimia monitor carries its own configuration: what to request, what counts as a failure, how often and from where to check, and who is alerted. New monitors all start on the same defaults. The editor marks every setting you change, so anyone opening a monitor can see which settings are off their defaults.

01Can Uptimia monitor an endpoint that needs authentication?+
Yes. HTTP basic auth is stored with the monitor and masked after saving. Token and key schemes go in custom request headers. For sign-in forms and session flows, transaction monitoring logs in like a real user, and an API monitor can post credentials in step 1 and reuse the returned token in later steps.
02What does a new monitor start with?+
HTTPS GET, every 5 minutes, a 5‑minute alert delay, 3 backup checks, all checkpoints (Europe for accounts registered in the EU), redirects followed, a 30‑second timeout, 4xx and 5xx responses plus connection, timeout, unknown, redirect, keyword and SSL errors reported, and the recovery notice on. You can change any of them per monitor, and the Advanced tab marks every card that is off its default.
03Which HTTP methods and timeouts can I set?+
GET, POST and HEAD. POST takes a raw body you write, and any method can carry custom headers. The timeout runs from 1 to 100 seconds, 30 by default. It is how long a checkpoint waits before a hanging request counts as a failed check. Keep it short for endpoints that must answer quickly; heavy admin pages may need a longer one.
04Can I allowlist the checkpoint IPs in my firewall?+
Yes. The full list is published in two formats: one IP per line, and the same list as lines for an nginx geo block. Both are generated from the checkpoint list itself, so they stay current when a location is added or re-addressed. Admit the whole list, IPv4 and IPv6. A hand-picked location set does not narrow the source addresses: confirmation checks can come from other locations, and some cities are served from a nearby checkpoint's server. The checkpoint IP list →
05Can one uptime monitor check a mail server or a port?+
Yes. An uptime monitor has three sub-types. Web checks a URL. Email checks SMTP, POP3 or IMAP on the port you set, with STARTTLS or SSL/TLS, and can expect a string in the server's greeting. Network opens a TCP connection, pings the host, resolves a name or sends a UDP packet, with an optional string to send and a string to expect back. Every sub-type keeps the same frequency, confirmation, recipients and locations.
06What happens to a monitor during a maintenance window?+
Its checks pause and its alerts are muted for exactly the window, then both resume on their own. Windows run once or repeat daily, weekly or monthly, cover one monitor, a group, a whole type or everything, and can be ended early from the dashboard. A window can also appear on your status page as planned maintenance.
07Can I stop alerts at night without pausing the monitor?+
Yes. You set it on the recipient, so the monitor keeps running. Each member and team carries alerting hours and blocked days, read in that person's own timezone. An alert that lands outside the window is held and delivered when the window opens, and the incident still opens and closes on time in the control panel.
08Can I copy a configuration to more monitors?+
Yes. Save a monitor as a template, up to 100 per account, and the next one needs only a name and a target. You can also clone an existing monitor, which carries its settings over, or paste a list of URLs into bulk create, where one set of settings applies to every row. Fleet-wide changes such as pause, resume, delete and maintenance run as bulk actions across a selection.
09Do all monitor types have the same settings?+
Every type keeps its name, recipients, groups, notes, maintenance windows and alert‑email theme. HTTP settings, the timeout, error classes and the unique ID belong to uptime monitors. SSL and domain monitors use an expiry threshold, page speed uses a load threshold, API monitors set method, headers, body and assertions per step, heartbeats use a schedule and a grace period, and real user monitoring is passive. Outage confirmation counts a little differently per type: website, SSL, speed and transaction monitors count backup checks from other locations, so 3 means three more locations, while API monitors count the first run as well, so 3 means the first location plus 2 others. A DNS monitor set to 1 alerts without a re‑check.
10Which settings depend on my plan?+
The check interval, branding and how many monitors you can run. 30‑second checks start on Professional, and so do branding themes that put your logo and colours on alert e‑mails and white-label your reports. The free plan checks 1 website every 5 minutes. Request settings, keywords, the timeout, confirmation, the alert delay and notes work the same on every plan, and the 30-day trial includes all of it.

Configure a monitor for your endpoint

Start on the defaults, then set the request, failure rules, timing and recipients your setup needs.

Start free trial →
30-day free trial Settings per monitor Changes marked in the editor No credit card
A new monitor starts on the defaults, and you can change any setting later.