Site health monitoring
Hourly uptime and certificate checks for every managed website, and what happens when a site goes down.
What is checked
Site health starts automatically when a website is created, and moves to the client's domain when it goes live. Every hour it requests the site and records the result.
| Result | Meaning |
|---|---|
| Up | The site answered normally within 5 seconds. |
| Degraded | The site answered with a client error (4xx) or took longer than 5 seconds. |
| Down | The site answered with a server error (5xx) or did not answer within 15 seconds. |
Open the project's AI tools tab and choose Site health to see the current status, uptime over the last 7 days (degraded checks count as not up), the median response time, how many days the SSL certificate has before it renews, and recent checks. Clients see the same view, read-only.
When a site goes down
- An urgent issue opens on the project, titled “Site health: site down” with the reason.
- The project Lead gets an email, “Site health: [project] is down”.
- If Slack or webhooks are connected, they get a site.down event.
- Emails and Slack alerts are sent at most once every 6 hours per site, and a second issue is not opened while one is still open.
When a certificate has 14 days or fewer left, the activity log notes it. Certificates for domains connected through ModelMerge renew automatically.
Pause or change the address
In Site health, set the Site URL and select Save settings. Use Pause agent and Resume agent to stop and restart checks, for example during a planned migration.
Still stuck? Email hello@modelmerge.ai with your agency name, the client and page involved, and the exact message you see.