Uptime, plainly stated.
Northstar checks your endpoints from five regions and turns the results into a public status page your customers can read in ten seconds — including on the bad days.
Northstar is a fictional product — this page is a design exercise in restraint.
api.example.com
OperationalSample data, generated on this page — the uptime figure above is computed from these 90 bars, not typed in.
Product
Three things, done carefully
Monitoring tools grow dashboards the way ships grow barnacles. Northstar does the three jobs a status product owes you, and stops.
Checks every 30 seconds
HTTP, TCP, and ping checks from Virginia, Frankfurt, Singapore, Sydney, and São Paulo. Every check is logged with region, latency, and response code — the raw rows are yours to export.
A status page on your domain
status.yourcompany.com, styled to your brand, readable without a login. Ninety days of history by default, with the incident log attached to the graph it explains.
Incidents without theater
Open an incident in one field, post updates as they happen, close it with a timestamp. Subscribers get email or webhook notices — written by you, not generated apologies.
Setup
Running before your coffee is
Point us at an endpoint
Paste a URL or an IP and port. Northstar starts checking within one minute and backfills nothing — history you see is history that happened.
Set your thresholds
Choose the regions that matter to you and how many consecutive failures count as down. The defaults are two failures from two regions.
Publish the page
Connect your domain, pick a logo, and share the link. Subscribers manage their own notifications from the page itself.
Pricing
One number per plan
Company
$79 / month
- 200 monitors
- Unlimited status pages
- SSO and audit log
- Private, gated pages
Billed monthly, per workspace. Every plan includes every check region.
FAQ
The questions that decide it
What exactly counts as downtime?
An endpoint is marked down after two consecutive failed checks from at least two different regions — one flaky route through one backbone shouldn't page you at 3 a.m. Both thresholds are adjustable per monitor. The downtime window runs from the first failed check to the next successful one, and that raw math is what your public uptime percentage is computed from.
Do slow responses count, or only errors?
You can set a latency budget per monitor — say, 2 seconds. Responses over budget mark the check "degraded," which shows on the status page as a distinct state but does not count against uptime unless you configure it to. Degraded-but-up is information your customers deserve; hiding it in "operational" is how status pages lose trust.
Can the status page live on our domain?
Yes — CNAME status.yourcompany.com to Northstar and the certificate is provisioned automatically. Your logo, your colors, no Northstar branding beyond a line in the footer. On the Company plan pages can also be private, gated by password or your SSO.
What happens when I hit my monitor limit?
Nothing dramatic: existing monitors keep running and alerting exactly as configured. You just can't add new ones until you remove one or move up a plan. There are no overage charges anywhere in the product.
Can I export my data if I leave?
Every check ever run — timestamp, region, latency, response code — exports as CSV or JSON from settings, on any plan, including after you cancel; exports stay available for 30 days. Uptime history is an argument you make to your customers, and it should remain yours.