docs/user/observe/uptime.md
title: Uptime description: See whether visitors can reach each service, check it after a deploy, and read its incidents.
Ployz checks every service with a public address from outside your servers, the way a visitor reaches it. When the check can't get through, the service shows a red ⚠ Unreachable and an incident opens.
See a service's uptime
- Click the service on the canvas.
- Open its Monitoring tab.
The status line reads Up, Unreachable or Checking, with the share of checks that passed. Response time and the history strip cover the last hour. Pick a quick range from 5 minutes to 30 days, or type a From and Until. Click an incident to see what each check got back. Below them come the service's metrics, on the same time axis.
Check right after a deploy
Ployz checks the services a deployment applied as soon as it succeeds, so a deploy that broke the address shows Unreachable within seconds. Between deploys, it checks every 5 minutes.
To check now, click Check now on the Monitoring tab. The button reads Checking until the result is in, and the status line moves with it.
From the CLI or an agent
On a Ployz Cloud backend, ployz uptime show prints what the Monitoring tab shows: the
status, the share of checks that passed, what each region's latest check got back, its last
error, and the incidents in the range.
| 1 | ployz uptime show web |
| 2 | ployz uptime show web --project shop --env staging --from 6h |
--project and --env default to PLOYZ_PROJECT and PLOYZ_ENV, then to where the
directory is linked with ployz link, else to your default Project and its default Environment.
A Project other than the linked one, typed or from PLOYZ_PROJECT, leaves the linked Environment
out, so you get that Project's default Environment unless you name one.
--from and --to take an offset back from now such as 15m, 6h or 7d, or a UTC instant.
The range is the last hour when you leave them out.
ployz uptime check does what Check now does, waits for the check to answer, and prints
what each region got back. It changes no setting. If the check hasn't answered within about 10 seconds, it says
so, and ployz uptime show shows the result once it lands.
| 1 | ployz uptime check web |
The sidebar and ployz mcp offer both as uptime_show and uptime_check, so an agent can see
that a service answers 500 and check it again after a fix. A service with no public address
answers not_checked.
Good to know
- What gets checked. The service's generated domain, or its first custom domain when it has
none, at its health check path, or
/without one. At the path, any answer below 400 counts as up. At/, any answer below 500 counts, so a 404 still shows the service is up. A service with no public address isn't checked, so its Monitoring tab starts at its metrics. - A 4xx points at the path. If the check has never passed and your app answers its health check path with a 4xx, the Monitoring tab shows the Healthcheck path field with what the app answered. Change it there, then deploy. The check uses the new path from then on.
- A failed deployment isn't checked at once. The regular check still runs, against what that deployment did apply.
- Uptime counts every check. The ones after a deploy and from Check now count toward the percentage and the charts, like the regular ones.