Product updates

What's new

• Alerts can go to more than one place. Destinations live in Workspace settings: a Slack incoming webhook or any HTTP endpoint, each with a name, an address and a switch. Send test posts one real message the way a real alert would and tells you what the far end answered. The address is treated as a credential — shown shortened, never echoed back into the edit form.
• Your existing webhook keeps working. The Webhook URL you had is now a destination named Default, switched on and marked as the workspace default. Nothing to set up, and nothing stops sending.
• Each subject decides who it tells, and about what. An Alert delivery block on the subject's page picks destinations and the kinds they hear — problems and reminders, no data, recoveries. The marketing site telling nobody no longer stops the payments API paging the on-call. A subject with rows of its own replaces the workspace default rather than adding to it, and the block says plainly which of the two is in force.
• Transformations shape what a destination receives. Name a template and that destination gets a PagerDuty event, your own JSON, or a line of text instead of the built-in message, with {{ field }} placeholders over a published vocabulary. A template is stored only once it renders a problem, a no-data and a recovery — a broken edit leaves the working one exactly as it was — and JSON values are escaped as they are substituted, so a quoted error message cannot break the document. Preview renders all three without saving.
• The delivery log says what was actually sent. Payload, status and time per destination for the last 30 days, with an Undelivered filter for the question you usually arrive with. A record keeps the destination's name as it was at the attempt, so a later rename still reads correctly. It is a read — nothing on the page resends.
• All of it is scriptable over the API and from Python SDK 1.8.0 on PyPI, which also adds Custom signal and rule authoring. Reading destinations and the log needs workspace edit rights — they carry webhook addresses and sent bodies.

What it costs

• Nothing changes. Destinations, alert delivery, transformations and the delivery log are part of alerting — they add no usage and no charge, however many destinations a workspace keeps. See Pricing.

What's new

• Say you have seen it. Acknowledge an open incident and the timeline records who did it and when, so the next person to look knows somebody is already on it. It changes nothing else: the verdict, the evidence and the recovery carry on exactly as before. Monitoring shows a small mark on the row, with the name behind it.
• A problem nobody has answered says so again. A confirmed problem now repeats every four hours, through the same webhook as the first alert, until somebody acknowledges it or it recovers. Acknowledging stops the reminders and recovery cancels them, and nothing is replayed after a quiet period.
• Maintenance windows. Before a deploy that will look like an outage, set a window on that subject — choose days and hours — and its problem alerts and reminders wait until it ends. Recoveries are still sent, monitoring and the timeline keep recording, and Monitoring marks the row while it runs. Change the end or finish early at any time.
• One page per subject. Opening a subject now shows Incident controls — acknowledge and maintenance — then its signals with what each one last received, then its rules, then the timeline. Adding a signal or a rule sits with the list it adds to, and opening a signal shows what it is, how to send data to it and its full history on one page.
• Acknowledge incidents and manage maintenance windows over the API, and from Python SDK 1.7.0 on PyPI. Website and Custom monitoring keep their own endpoints.
• A database outage no longer stops the server. It waits and reconnects on its own, and picks up where it left off — no restart needed.

What it costs

• Nothing changes. Acknowledgement, maintenance windows and reminders are part of monitoring — they add no usage and no charge, and reminders go to the webhook you already have. See Pricing.

What's new

• Monitor jobs and scripts with Add Custom check. This creates a Custom subject for your own signals, separate from your Website checks.
• Add a heartbeat for regular reports: if it stops, you get missing data. Use an event for occasional reports: silence is normal, and the last reading still stands.
• Send readings through the API or Python SDK 1.6.0. Report OK or problem, with an optional number, error message and labels.
• Build rules from signals and other rules on the same Custom subject. Choose what counts as a problem, how many inputs must agree, and when to confirm or close an incident.

What it costs

• Custom observations count toward hosted usage too. Each accepted heartbeat or event uses one observation unit, just like a website probe. Rejected observations don't count; retries that replace an existing observation don't add usage. See Pricing.
• Self-hosted Uptimer does not meter or bill observations.

What's new

• Monitoring is organized around your websites. Each row shows whether that site is OK, still confirming a problem, recovering, or missing data — not just up or down.
• Open a website to see a timeline of what happened: when trouble started, what each location saw, and how status changed over time.
• Add or edit a website monitor in one form — setup stays simple while the product keeps the full picture behind the scenes.
• Slack alerts link straight to the website that needs attention, so you land on the story instead of hunting through menus.
• New API keys show the secret once when created; copy it then — it will not appear again.

Features

• A monitor's status is now decided across all of its locations, with a new "Locations Required to Fail" choice — majority (default) or at least one.
• History is now a timeline of incidents — pending, problem, no data, recovering, closed — showing which locations failed and what they reported.
• Slack alerts were rewritten. Each one now says how many of your locations are failing and which, quotes the error the site returned, and links straight to the timeline. Recovery messages say how long you were down.
• The monitor list shows Pending while a problem is being confirmed and Recovering while a site is coming back, so you can see what is happening before an alert arrives.
• Monitoring requests identify themselves as Uptimer/1.4.0 instead of Go-http-client/1.1, so you can filter them out of your analytics.
• Signing out now ends the session at your identity provider too self-hosted OIDC

Features

• Assign regions to a rule when creating or updating it over the REST API — a new "regions" field on the rule payload, matched by region name.
• Python SDK v0.4.0: set regions on rule create/update published to PyPI

Features

• Added Instant Check feature with sharable results.
• Introduced Singapore region (SG).

Features

• CSS rework with minification and class mapping CSS/JS reduced ~50%
• HTML size reduced through CSS class minification

Fixes

• Fixed paging issue on the rule availability history page.
• UI error messages updated for clarity.

• Version number displayed alongside project name
• Improved OIDC-to-user mapping (security fix)
• Fixed header spacing issue on the rules list page

SaaS updates

• Launched public API Officially launched our public API 🚀
• Redesigned pages for improved clarity and navigation Rolled out fully redesigned pages for a smoother experience

Python SDK v0.3.0

• Self-hosted client now requires a base_url
• Introduced cloud client for integration
• Updated documentation for clarity

Python SDK v0.2.0

• First public Python SDK version
• Added full rule management (create, update, delete, list)
• New APIs for regions, rules, and workspaces

Uptimer (self-hosted) v1.0.0

• New APIs: rules (CRUD), version, regions, workspaces
• UI supports API key and token management
• Core refactor: app initialization, services, handlers, tests Major core refactor: app initialization, services, handlers, tests
• Refreshed UI
• Optimized performance Fewer SQL queries, parallel API tests

Documentation

• Updated documentation for Uptimer v1.0.0

• Documentation launched with initial version of Uptimer (v0.2.0)