← Back to Index
Methodology
How Checks Work
We send HTTP HEAD requests from 15 probe servers worldwide, every 5 minutes. Each request has a 10-second timeout and uses a custom User-Agent identifying the bot.
If a HEAD request returns a non-2xx response (except codes that indicate the server is responding (e.g., 400, 401, 404, 405, 406, 429)), we retry with a GET request.
Status Classification
UPMore than 50% of probes return 2xx AND average response time is under 500ms.
DEGRADEDMore than 50% of probes return 2xx BUT average response time is 500-1500ms OR more than 20% of probes exceed 1500ms.
DOWNMore than 50% of probes return 5xx, timeout, or fail to connect.
PARTIALSome probes return 2xx, some return errors. Less than 50% failure rate.
BLOCKEDAll probes receive HTTP 403. This indicates WAF blocking, not a real outage.
NO_DATANo check results in the last 10 minutes.
Probe Locations
| City | Country |
|---|---|
| New York | US |
| Chicago | US |
| Los Angeles | US |
| Toronto | CA |
| São Paulo | BR |
| London | GB |
| Frankfurt | DE |
| Amsterdam | NL |
| Warsaw | PL |
| Stockholm | SE |
| Tokyo | JP |
| Singapore | SG |
| Mumbai | IN |
| Sydney | AU |
| Johannesburg | ZA |
Known Limitations
- WAF blocking: Some sites block automated requests. These are shown as BLOCKED, not DOWN.
- Rate limiting: Some APIs may rate-limit our bot, resulting in 429 responses.
- Regional routing: CDN-backed sites may route to different origin servers from different probes.
Request Site Removal
If you are a site owner and would like your site removed from our monitoring index, please contact us at hello@parlon.io.