Uptime checks and incidents
How LopeBase checks your sites every 15 minutes, watches the services behind them, and reviews outages for you. Set up a check and read what it shows.
Uptime checks every site you add every 15 minutes, with no account or key needed, and keeps a year of history. On RunUptime the Overview tab opens with one line saying whether everything is up. Anything down or degraded is pinned right under it with buttons to open the property, post a status note, or see the outage. Below that are your uptime over 30 days, outages this month, and average response time across sites, then a card for each site with problems first. The Outages, Monitors, and Status pages tabs hold the rest, and a tab with nothing in it is left out.
Add a check
A new property has no check until you add one. Properties with no check yet are listed in a short panel at the bottom of the Overview tab, each with an Add check button.
-
Add the address
Give the site's own address, or a health address if it has one. LopeBase fetches it every 15 minutes and marks it down if it does not answer, answers slowly, or returns an error.
Add an uptime check -
Or get it from the adapter
If the property already runs the LopeBase Adapter, connecting it adds its health address as the uptime check automatically, unless that property already has one.
Watching the services behind the site
If your app runs the LopeBase Adapter, point the check at its health address (/api/lopebase/health by default) instead of the home page. The adapter answers with each service you declared there, like your database, a queue, or a second API, and each one shows as its own row on Uptime.
Each service is checked with the app's own credentials, like a select 1 or a Stripe balance.retrieve, so it proves the integration works rather than just that the vendor's servers are up.
A service you mark non-critical shows the whole site as degraded if it fails, not down. A critical service failing answers 503 and the site shows as down. Degraded still counts as up in your uptime percentages, but it shows amber and labeled "Degraded" apart from a real outage everywhere it appears: the day strip, a status page's badge and 90-day bars, and the alert you get.
Reading the numbers
Each site's card shows its 30-day uptime and average response time up front, then a 30-day strip colored per day: green for a clean day, orange for degraded, red for any failed check. Hover a day for its tally. The line under the strip has 24 hours and 7 days from individual checks, plus 90 days and 12 months from daily totals once you have history. A window with no checks yet in it says "Collecting" rather than a number. The uptime figure at the top of the page is the plain average of every site's 30-day figure.
Outages and the incident review
On the Outages tab, every outage in the last 30 days lists when it started, how long it lasted, and what the check saw, with the response-time chart for the last 24 hours under it. Below that, Review incidents reads those 30 days together with your top errors when Sentry, Bugsnag, Rollbar, Honeybadger, or PostHog's error tracking is connected, and says what went wrong and what to fix first.
After real downtime, the review drafts a status note for your customers. If you run a status page for that property, a button posts the note straight to it, in the wording the review drafted.
Error trackers and Better Stack
Uptime checks sit next to error trackers under Monitoring on ManageConnections. Connecting Sentry, Bugsnag, Rollbar, or Honeybadger adds that property's open errors and daily error counts to its timeline and to the incident review, read-only.
If you already run monitors in Better Stack, connect it with one team API token and LopeBase reads the monitors you pick. They show on the Uptime page's Monitors tab, with each one's status and 30-day availability, and a monitor that is down counts in the summary at the top like a failed check and is pinned there with a button to see it. It also shows under Needs attention on Overview and alerts you right away. Better Stack monitors sit beside LopeBase's own 15-minute checks; a property can have both.
If something goes wrong
Something else? Contact us.
Why does Uptime say "Collecting" instead of a percentage?
A window needs at least one check inside it. A brand-new check has none yet for the shorter windows and takes 90 or 365 days to fill the longer ones. The live status at the top is accurate right away.
LopeBase says a site is down but it loads fine for me
Check the address you gave it: a page behind a login, a redirect to a different host, or a firewall that blocks LopeBase's requests all read as down. The outage's "What we saw" column shows the HTTP status or error LopeBase got.
Why does it say degraded instead of down?
One of the services your health address lists failed, but you marked it non-critical, or the site itself still answered while a critical service failed partway. The error message names which service.
Can I watch something other than a web address?
Only through a health address that answers over HTTP. There is no check for a database or queue you cannot reach from the public internet, except through your own LopeBase Adapter's health endpoint.
I don't want LopeBase pinging a vendor's status page
It never does. A service check always uses the app's own credentials and a cheap authenticated call, never a vendor's status page or home address.
Why did I get an alert for something that recovered in a minute?
A short blip still shows on the day strip and in the outage list, but alerts wait until something has failed for about an hour before emailing or pushing, so a brief hiccup does not page you. See Alerts.