Create a Monitor
Set up website, port, SSL, DNS, domain and server health checks that run from outside your network, try a safe fix first, and alert you if it does not work.
Before you start: A public address to check, or a connected server for a server health check.
- 1
Open Monitoring and choose New monitor
Monitoring is under Operations. Checks run from the platform, not from your server, so they can detect a server that is completely unreachable.
- 2
Choose what to check
Website or API requests a page and checks the response. Port checks that a service accepts connections. SSL certificate watches an expiry date. DNS record checks that a record still contains the expected value. Domain expiry checks the registration with the registry. Server health watches disk, memory and load on a connected server.
- 3
Set the expected result and interval
A website check can require a phrase on the page, which catches an application that fails while the server still responds. The shortest interval is one minute. A failure must repeat before it counts, so a single blip does not trigger an alert.
- 4
Attach a server for Auto-Fix
With a server selected, Try to fix this automatically lets the platform restart stopped services, reload PHP-FPM or the web server, or free disk space before alerting you. It never reboots, deletes data or changes firewall rules, and it stops after three attempts in a day.
- 5
Choose where alerts go
Where to send alerts takes email addresses and webhooks, added once and used by every monitor. Without one, a monitor detects outages but notifies nobody.
- 6
Review the summary
Up, Down, Not checked recently and Average uptime over the last 24 hours, the same four figures shown on your dashboard. Not checked recently counts monitors whose last check is older than three of their intervals, so their status may be out of date. The average leaves out SSL and domain expiry monitors, which count days rather than uptime, and only includes monitors with at least five checks in the day.