Your Status Page Shouldn't Go Down With Your Server
When something breaks, your customers want one answer: is it you, or is it me? If they can't find it in a minute, they open a support ticket. Here is how a status page does that job, why it has to live outside your own infrastructure, and how OnCallAlert gives you one with every plan.
The problem with a status page on your own servers
A lot of teams put the status page on the same cluster as the product. It is easy, and it works fine on a good day.
On a bad day the cluster is down, so the page is down too. Customers see an error where your explanation should be. You have explained nothing, and you look like you don't know what is happening.
The page has to be reachable exactly when your own stuff is not. That is the whole point of it.
The second problem: someone has to update it
Picture the incident. You are reading logs, someone is rolling back a deploy, and somewhere in the middle you are supposed to open a page and change a dot from green to red. Most of the time that does not happen until it is almost over. Then the page says “all good” while customers are already angry.
A status page that depends on a human remembering is a status page that is wrong at the worst moment.
How OnCallAlert does it
Every plan, including Core, comes with one public status page at yourname.oncallalert.dev. Three things make it useful.
1. It is hosted outside your servers
Your OnCallAlert server sends the page to our hosting service whenever it changes. Visitors are served from there, not from your machines. If your server is having a terrible afternoon, the page still loads.
One honest limit: a server that is completely down cannot send new updates. The page keeps showing the last thing it knew. For that case, use a manual status or write an incident by hand, and pair it with the dead-man's switch so you hear about the outage first.
2. It follows your alerts on its own
Each component on the page can be linked to a service you already have in OnCallAlert. After that, the status updates itself:
| What is happening on the linked service | What customers see |
|---|---|
| An alert nobody has acknowledged, or a failed heartbeat monitor | Outage |
| Only acknowledged alerts | Degraded |
| No open alerts | Operational |
Changes show up within a minute or two. The on-call engineer acknowledges the alert and the page already says something is wrong. Nobody had to switch tabs.
You stay in control. Any component can be set by hand, and it keeps that status until you switch it back to Automatic. That is handy for planned maintenance, or for a service that raises noisy alerts customers never feel.
Tip: every open alert on a linked service changes the public page. Link only services where an alert really means a customer-facing problem. Set the rest by hand.
3. You write the incident updates yourself
Automatic status tells people that something is wrong. Incidents tell them what you are doing about it. You write those yourself: a title, the impact, a first message, then updates as things move along, and you mark it Resolved at the end. The page shows your 20 most recent incidents, so people can see you handle problems properly.
There is also a 30-day history, one bar per day for each component. A day's uptime is the share of that day the component was not in Outage. Time is only counted while your server is running, so a period when your own server was down is not recorded as an outage. Keep that in mind when you read the bars.
Set it up in about five minutes
- Open Admin → Status Page (admins only).
- Pick an address (3 to 40 lowercase letters, digits or hyphens, starting with a letter). The form tells you right away if it is taken.
- Add your components and link them to services.
- Turn on Published and save.
The page appears within a minute or two. Your server needs a valid license key and outbound HTTPS access
to statuspages.oncallalert.dev. If you want to take the page down later, switch Published off or
delete it, and your server asks the hosting service to remove it.
Want status.yourcompany.com?
Team, Business and Unlimited can use their own subdomain as a $5 per month add-on. You add one
CNAME record at your DNS provider, press “Check again”, and it goes from Pending to Active, usually
within a few minutes. The free oncallalert.dev address keeps working too. A bare domain like
yourcompany.com is not supported, only subdomains.
Things to know before you publish
- Everything on the page is public. Don't put personal data or internal hostnames in it.
- Machines can read it. Every published page has a
/status.jsonat its own address, handy for monitoring tools and AI assistants. - It is one page per plan. If you run several products, use components to split them.
What it costs
Nothing extra. One status page is part of every plan: Core ($699/year), Team ($1,499/year), Business ($4,999/year) and Unlimited ($9,999/year). Plans are priced by team size, not per seat, and viewers are free. Many hosted status page products charge per page, per subscriber or per component on top of your on-call tool. Here it is already in the price.
Common questions
Is the status page really included in every plan?
Yes. Core, Team, Business and Unlimited each get one public status page at yourname.oncallalert.dev at no extra cost.
Does the status page stay up if my own server is down?
Yes. Your server sends the page to our hosting service whenever it changes, and visitors are served from there, not from your server. One limit: while your server is down it cannot send new updates, so use a manual status or a hand-written incident for that period.
How does a component decide its status?
Link it to a service. An unacknowledged alert or a failed heartbeat monitor shows Outage, only acknowledged alerts show Degraded, and no open alerts show Operational. You can also set any status by hand.
Can I use status.mycompany.com?
Yes, on Team and up, as a $5 per month add-on. You add a CNAME record at your DNS provider. A bare domain such as mycompany.com is not supported.
Is there a machine-readable version?
Yes. Every published page has a /status.json at its own address.
Want a status page that updates itself?
Run OnCallAlert on your own server and get a hosted public status page with it.
Sources and notes
- Setup steps, status rules and custom domain details: OnCallAlert status page documentation.
- Plans and the $5/month custom domain add-on: pricing.