HTTP Monitors
Monitor any HTTP or HTTPS endpoint. Check availability, response time, SSL certificates, and response content.
How it works
HowlOps sends HTTP requests to your endpoint at the configured interval from multiple locations. If the check fails (wrong status, timeout, SSL error, keyword missing), an incident is triggered and you're notified via your configured alert channels.
Live monitor: api.yourcompany.com/health
UP
· Last response: 127ms · Check interval: 60s
30-day view · 99.3% uptime
Configuration
| Field | Description | Required |
|---|---|---|
| Name | Human-readable monitor name (e.g. "API Health") | Yes |
| URL | Full URL including protocol: https://api.example.com/health | Yes |
| Check interval | How often to check: 120s on Free, 60s on S, 30s on M, 15s on L | Yes |
| HTTP method | GET, POST, PUT, PATCH, HEAD | No |
| Request headers | Custom headers: Authorization, Content-Type, etc. | No |
| Request body | JSON/text body for POST/PUT requests | No |
| Expected status | HTTP status code to consider as success (default: 2xx) | No |
| Keyword check | Text that must appear in the response body | No |
| SSL check | Alert if SSL certificate expires within N days | No |
| Timeout | Max response time before marking as failed (default: 30s) | No |
| Alert threshold | How many consecutive failures before triggering alert (default: 2) | No |
Authentication
To monitor authenticated endpoints, add headers in the Request Headers field:
X-API-Key: your-key
POST / PUT requests
Select POST as the HTTP method, set Content-Type: application/json in headers, and add your body:
Keyword monitoring
Use keyword checks when HTTP 200 isn't enough. For example, if your site serves a cached error page with a 200 status, keyword monitoring catches it:
- Keyword
"OK"must be present → alert if response is{"status": "error"} - Negate: keyword must not be present (e.g.
"maintenance")
SSL Certificate Monitoring
Enable SSL monitoring to receive alerts before your certificate expires. Configure warning thresholds:
- Warning: 30 days before expiry
- Critical: 7 days before expiry
Alert threshold
The alert threshold prevents false alarms from transient network hiccups. With threshold = 2, HowlOps waits for 2 consecutive failures before triggering an incident. This means you won't get paged at 3am for a single 500ms blip.
Degraded status
A monitor can show a Degraded state when it is technically up but performing poorly. This happens when:
- Response time exceeds the configured slow threshold (e.g., > 2000 ms)
- An SSL certificate is within the expiry warning window but not yet expired
Degraded is an intermediate state between Up and Down. A separate alert channel or lower-priority notification can be configured for degraded events so your team is aware without being paged immediately.
Automatic incident correlation
HowlOps automatically groups related incidents when multiple monitors fail near-simultaneously. Instead of flooding your team with individual alerts for a shared outage, a single correlated incident is raised.
- If your API, database, and frontend all fail within a short window, one correlated incident is raised.
- If only one monitor fails independently, it gets its own incident as usual.
Prober regions
HowlOps runs checks from geographically distributed prober nodes. Each monitor can be assigned to one or more regions:
- Nuremberg: Germany (EU)
- Helsinki: Finland (EU)
- N. California: United States
- Singapore: Asia-Pacific
When multiple regions are assigned (on any paid plan), checks run in parallel and results are aggregated. A failure in all regions = Down. A failure in some regions = Partial (regional outage). Degraded is a separate state that indicates slow response times.
See Multi-Region Checks for full configuration details.