Key Takeaways (BLUF)
Bottom Line Up Front: Self-hosted WordPress monitoring plugins — those that run monitoring code on your own server — cannot detect downtime when your server crashes, because the plugin stops running at the same time the server does. External cloud monitoring checks your site from independent infrastructure and sends alerts even when your server is completely offline. For genuine WordPress uptime monitoring, external is the only architecture that works.
What Makes a WordPress Monitoring Approach "Self-Hosted" vs "External"?
The distinction is about where the monitoring code runs.
Self-hosted monitoring: A WordPress plugin installs code on your server. That code runs periodic checks — either checking your own site from the inside (pinging localhost) or checking third-party URLs. The monitoring depends on your WordPress installation being operational to function.
External cloud monitoring: An independent service, running on completely separate infrastructure in one or more data centres globally, makes HTTP requests to your WordPress site URL from the outside. This infrastructure is entirely separate from your hosting environment — a different company, different servers, different networks.
The Decisive Comparison: What Happens During Real Downtime
The most important test for any monitoring approach is whether it works during an actual outage. Here is what happens to each approach during the most common WordPress downtime scenarios.
| Downtime Scenario | Self-Hosted Plugin | External Cloud (WakeMetric) |
|---|---|---|
| Server runs out of memory (OOM kill) | PHP stops. Plugin cannot run. No alert. | External check detects error response. Alert sent. |
| PHP fatal error — site crashes | Plugin cannot execute. No alert. | External check detects 500 error. Alert sent. |
| Database server crash | Plugin cannot query DB to schedule or send alerts. No alert. | External check detects "Error Establishing a DB Connection" response. Alert sent. |
| Web server (Nginx/Apache) crash | Plugin cannot be served. No alert. | External check gets connection refused. Alert sent. |
| Hosting provider network outage | Entire server unreachable. Plugin cannot send alerts. No alert. | External nodes independently detect unreachability. Alert sent. |
| WordPress admin accessible but front end broken | Plugin may detect if coded to check front-end URL | External check always checks the configured public URL. Alert sent. |
| Site returns 503 during maintenance | Plugin may detect if designed to catch this | External check detects 503. Alert sent. |
Use Cases Where Self-Hosted Monitoring Makes Sense
Self-hosted monitoring plugins are not universally useless — they have legitimate use cases. Understanding when each approach is appropriate helps you choose correctly.
- Monitoring third-party services FROM your server: If you want to know whether your payment gateway, email API, or external CDN is responding, a self-hosted plugin can make those outbound checks from your server's perspective
- WooCommerce-specific health checks: Plugins like Titanium Site Watch run application-level checks (cart function, checkout process, database table integrity) that require running inside WordPress to execute properly
- Dead-man switch heartbeats: Your WordPress application pings an external endpoint — the reverse of self-hosted monitoring, where your app is the sender and an external service is the receiver
Why WakeMetric Chooses External Cloud Architecture
WakeMetric's uptime monitoring uses external cloud infrastructure in five global regions because the only monitoring architecture that reliably alerts you to WordPress site downtime is one that does not depend on your WordPress site being online.
The WakeMetric WordPress plugin is a lightweight connector — it sets up your monitor and displays the widget. The actual checking happens externally. This design means your alerts work in every scenario, including the ones where your server is completely offline.
Frequently Asked Questions
Can a self-hosted plugin detect downtime if it checks via a remote API?
If a self-hosted plugin uses an external API or service to perform the actual check (rather than checking from the server itself), the monitoring can be external. However, the plugin still needs WordPress to be running to schedule and execute the API call. If your server crashes before the API call is made, the check is not performed. True external monitoring runs on a completely independent schedule that is not dependent on your WordPress installation in any way.
My hosting plan includes uptime monitoring. Is that external?
Hosting provider monitoring typically monitors the server infrastructure (CPU, memory, disk, network connectivity) from within their own data centre network. It does not monitor your WordPress application from a visitor's perspective. A monitoring tool that checks from inside the hosting network may still report your site as "up" when it is returning 500 errors to real visitors — because the hosting network's internal check may bypass the error at a different layer. External monitoring from a completely different provider's infrastructure is the most accurate proxy for the real visitor experience.
Conclusion
Architecture determines effectiveness in WordPress monitoring. A plugin running on the monitored server is like asking a guard to report whether the building they are in is on fire — they cannot report if the fire has already incapacitated them. External cloud monitoring from independent infrastructure is the only approach that guarantees alerts during actual outages. Install the free WakeMetric plugin and add genuinely external WordPress monitoring to your site today.