What we shipped in Uptimecheck, newest first: new features, improvements, fixes and security updates to monitoring, alerts and status pages.
We updated the web framework that app.uptimecheck.com runs on to a patched release. It closes three recently published vulnerabilities in that framework. Nothing changes in how the app works, and you don't need to do anything.
Every status page now has a history page at /s/<name>/history. It lists the incidents you wrote, with all their updates and impact, and the outages our checks detected, with their incident report.
Incidents are grouped by month in UTC, 20 to a page, and go back as far as your plan keeps history.
We checked every alert channel against the service's own API. A monitor name or message can no longer trigger @channel or @everyone mentions in Slack, Discord, Google Chat or Mattermost, and each service's length limits are respected.
A channel that fails for good, such as a deleted webhook, stops retrying at once and shows a clear reason, and a service asking us to slow down (Retry-After) is listened to. Webhook payloads now include the monitor's ID and the incident.
When you report an incident, you can now set Degraded performance, Partial outage or Major outage for each component it affects. The status page, badge, widget and summary.json show the worse of your choice and what our checks see, so a confirmed outage is never hidden.
The REST API and the MCP server take the same impacts.
Maintenance windows that haven't ended can now be edited, and so can recurring series. A window that is running keeps its start, and a series changes from its next window on.
Subscribers who were told about a window get a "Maintenance moved" notice with the new and the old time, in all 9 status page languages. The REST API can change windows too, and the MCP server can list and update them.
Choose a plan and monthly or yearly billing on our pricing page, and sign-up now remembers it. The app shows your choice when you sign up, and Billing preselects that plan and interval when you are ready to subscribe. The 14-day trial is the same as before.
A new monitor now offers to send a test alert until one of your channels has delivered one, and the getting-started checklist has a "Send yourself a test alert" step. It is the quickest way to know an alert will reach you before the first real outage does.
Alert emails now name the cause correctly, for example "answered with an error" for an HTTP 503, and expired certificates and domains say that they have expired. Each address gets its own message, so recipients don't see each other, and the unsubscribe link in a client report is signed for its recipient.
New workspaces get a few getting-started emails, which you can turn off under Account. Trial and billing notices always arrive.
A new free tool at /tools/uptime-calculator/ turns an SLA into allowed downtime per day, week, month, quarter and year, or turns your downtime into an uptime percentage. 99.9% allows 43m 49s of downtime a month. Every result has its own link to share, and it works without JavaScript.
Read more: What Does 99.9% Uptime Mean? SLA Downtime Explained
Every alert channel now shows whether its alerts arrived over the last 30 days. A failed alert is retried five times over about 40 minutes. After that the channel shows Not delivering with the reason: at the top of the overview, on the monitors that use it and in the weekly report.
Fix the channel and press Send test. A test that gets through clears the warning.
An incident that ended because a maintenance window started, or because someone paused, changed the address of or deleted a monitor that was down, used to close silently. Slack and email never heard it was over, and the PagerDuty incident stayed open.
Now every channel that got the down alert gets a notice that says why the incident closed, and PagerDuty resolves it.
Read more: How to connect PagerDuty, Google Chat, ntfy and more
The weekly email report now follows each person's time zone, set under Account › Email reports: Monday 00:00 to Monday 00:00, sent from 07:00 local time. A report that was missed is sent afterwards, and none is sent twice. The reports page and its CSV export count months in the same time zone.
Monitors and our free website checker refuse addresses that point into private networks or a cloud provider's metadata service. We closed a gap where one way of writing an IPv6 address could get past that check.
When our own monitoring had a gap, cron job and server monitors could alert as soon as it ended. They now pause during the gap and start their clock again when it ends.
Alerts also go out six at a time instead of one by one, so a single slow webhook can no longer hold up the rest.
The workspace overview now opens with one sentence about where things stand. Below it: Needs you now, with Acknowledge buttons; Worth a look, for things like a certificate about to expire; your services with response times against last week; and the week's uptime. The services you depend on are listed under Your providers and don't count toward your own uptime.
The menu is shorter too: Monitoring holds Overview, Monitors and Incidents, and Reports and Maintenance sit under Workspace.
Set up your first monitor in under a minute, or look around the live demo first.
14-day free trial · No credit card required
With your OK, Google Analytics and Microsoft Clarity show us which pages help and where visitors get stuck. They set cookies for up to 13 months. Without your OK, neither of them runs. You can change your mind at any time under Cookie settings at the bottom of the page. Privacy policy