Uptime

Monitoring client websites, what counts as down, and the incident history that proves it.

/admin/uptime — whether your clients' websites are answering, how quickly, and what happened the last time one stopped. Adding a website marked Live to a client already started this; the page is where you confirm it and manage the rest.

Reading needs uptime.view. Starting, pausing and stopping needs uptime.manage.

#The monitor list

The uptime page listing monitors by client, each with a state badge, response time and controls to check now, pause or stop. — Client websites that are not yet monitored appear with a Start monitoring button on their row.

Grouped by client, with the client name linking to the client record. Per monitor:

Check now — an immediate check, outside the schedule.
Pause / Resume — stop checking without losing history.
Stop — remove the monitor.

Client websites that are not yet monitored appear with a Start monitoring button on their row, so nothing has to be re-typed.

Stop discards the incident history Pausing keeps everything and stops checking. Stopping removes the monitor and the record of what it saw — which is the record of what your service actually delivered. If you might want that later, pause.

#Monitoring something that is not a client website

Add a site manually opens a short form: an address, an optional name, and a client — or No client — ours for your own site, a staging environment, or a landing page that lives outside the client's domain.

#What counts as down

A single failed check is not an outage. A site is reported DOWN only after two consecutive failures, and recovers after two consecutive successes. Between the first failure and the second, the badge shows the interim failure count beside it — so a wobble is visible without being announced.

Setting	Value
Check interval	5 minutes
Consecutive failures before DOWN	2
Consecutive successes before UP	2
Request timeout	15 seconds
Redirects followed	Up to 5
Individual check rows kept	7 days

Individual check rows are trimmed after a week by a maintenance job; the daily roll-ups and the incident records are kept, so long-run uptime figures survive the trim.

#One monitor

/admin/uptime/[id] — its recent checks, response times and incident history, with Check now and Pause / Resume, and a link to the client.

#Being told

An outage sends an email. Checks are performed by a sweep that runs every monitor whose interval has elapsed — it is a background job, so nobody has to have a tab open. See Background jobs.

#Limits

Monitors are bounded per client by the platform's uptime monitors per client ceiling. The number is set in the platform price book and can be excused for one agency; see Limits and cost.

#What the client sees

Uptime has no dedicated page in the client portal. Its value to the client is context: the SEO work makes more sense when you can say a ranking fell the week the site was down. That is also why the SEO Specialist role holds uptime.view and not uptime.manage — the history is evidence, and deciding what gets monitored is an account decision.