Everything we have shipped.

A site that only claims what exists

StatusOwl is a status page company, so this site was rebuilt to say that plainly — and every claim on it was checked against the code before it was published.

• Improvedstatusowl.net is now organised around the status page itself: what goes on it, how it updates itself when checks fail, and how it compares to the alternatives.
• ImprovedPlans are laddered by how many status pages you run, not by how many people read them. The number of people watching during an outage does not change your bill.
• NewA public roadmap listing what is not built yet — including status page subscribers and alerting on failing uptime checks — each with a plain statement of where it actually stands today.

Operator II, and history that gets pruned

A second-generation dashboard, plus two things that were quietly not working the way the settings said they were.

• ImprovedThe dashboard has a new instrument-style design. Layered surfaces, monospace readouts for every figure, and a clearer split between what a service is doing now and what it has been doing.
• FixedData retention now actually applies the window your plan includes. It had been running as a no-op, because no plan carried a retention setting at all, so old check data was never pruned. A one-off cleanup pass reclaimed roughly 250 MB of stored history.
• FixedEditing a monitor's configuration is now handled cleanly end to end. Results that were already in flight against the old configuration had nowhere to land and were retried until they failed; they are now discarded deliberately and quietly.

The last service off the old server

• ImprovedOur internal admin console moved onto Cloudflare Workers with a pooled database connection, finishing the move started in May. Nothing to do on your side — it is the last piece of StatusOwl that was still tied to a single machine.

Off one server, onto redundant infrastructure

StatusOwl used to run on a single dedicated server. The page that tells your users something is broken should not have a single point of failure of its own, so it no longer does.

• ImprovedStatus pages now run on more than one instance behind Cloudflare. Losing a machine no longer takes your page down at the exact moment people are visiting it.
• ImprovedThis site and the public API moved to Cloudflare Workers. The dashboard, status pages, the API, the check-result consumer and the scheduled jobs moved to Fly.io.
• NewThe public API enforces rate limits per key, so one key cannot exhaust the service for the rest of your organisation.
• ImprovedBoth databases moved to managed Postgres with continuous backups and point-in-time recovery covering the last 7 days.

A smaller platform

We retired several services left over from earlier versions of the architecture. Fewer moving parts sit between a check running and its result appearing on your page.

• RemovedRetired the legacy WebSocket gateway and the activity-stream service behind it. The dashboard refreshes over plain HTTP polling instead.
• RemovedRetired the remaining unused cloud services. A single message queue is all that now carries a check result from the agent that ran it to your status page.

Public status page refresh

A visual overhaul of the public page, and the end of the most annoying part of setting up a custom domain.

• NewLight and dark themes on public status pages, with a visitor-facing toggle. It follows the system preference by default and remembers what each visitor chose.
• ImprovedThe theme is applied before first paint, so a visitor on a dark page no longer gets a white flash on the way in.
• FixedCustom domain verification no longer needs a manual refresh after DNS propagates. The setup screen polls hostname and certificate status and clears each step as it completes.

Better Auth migration

Session handling moved to Better Auth: a smoother sign-in, and sign-outs that take effect immediately.

• NewAn active sessions list in user settings, with device labels. Sign out of one device individually, or all of the others at once.
• ImprovedSessions are opaque tokens checked against the database on every request rather than self-contained tokens, so revoking one takes effect immediately instead of when it happens to expire.
• SecuritySession cookies are signed with HMAC and verified on every API request.

SSL and domain monitoring, generally available

• NewCertificate monitoring: issuer, subject, valid-from and valid-to, and days remaining, with your own warning and critical thresholds.
• NewDomain registration expiry over WHOIS, including a warning when your nameservers change unexpectedly.
• ImprovedHTTP checks take a request method, a list of expected status codes, custom request headers, a request body, a follow-redirects toggle and their own timeout.

Audit logs and roles

• NewAudit logs record every incident, maintenance window and status page change, with who made it and when. Included on every plan, Free included.
• NewOwner, admin, billing and member roles per organisation, checked before every read and every write.