Multi-location Checks
When a site appears to be down, SiteBrief verifies the outage from 6 global locations before sending an alert — eliminating false positives.
The problem it solves
Network issues happen constantly — a temporary routing problem, an overloaded CDN node, or a brief glitch between our monitoring server and your client's host. Without multi-location checks, these transient issues trigger false alerts that wake you up at 3 AM for a site that was only "down" for 5 seconds from one specific network path.
Multi-location checks solve this by asking: "Is the site actually down, or is there a temporary network issue between us and the site?"
How it works
When the primary monitoring check fails, SiteBrief immediately triggers simultaneous checks from all 6 monitoring locations. The alert decision is based on how many locations confirm the outage:
| Locations reporting down | Action |
|---|---|
| 1 out of 6 (primary only) | No alert — likely a transient network issue |
| 2–3 out of 6 | Alert sent — partial outage or regional issue confirmed |
| 4+ out of 6 | Alert sent — confirmed global outage |
Monitoring locations
| Region | Location |
|---|---|
| 🇺🇸 North America East | Virginia, USA (AWS us-east-1) |
| 🇺🇸 North America West | Oregon, USA (AWS us-west-2) |
| 🇩🇪 Europe Central | Frankfurt, Germany (AWS eu-central-1) |
| 🇮🇪 Europe West | Dublin, Ireland (AWS eu-west-1) |
| 🇸🇬 Asia Pacific | Singapore (AWS ap-southeast-1) |
| 🇧🇷 South America | São Paulo, Brazil (AWS sa-east-1) |
Each location makes an independent HTTP request to your site. You can see the response time from each individual location in the Multi-location check panel on the site detail page.
Running on-demand location checks
You can trigger a manual multi-location check at any time from the site detail page. The panel shows the response time and status from each of the 6 locations — useful for:
- Diagnosing regional issues (e.g. site is slow in Asia but fast in Europe)
- Verifying a fix after an incident
- Understanding performance for a globally distributed audience
- Testing CDN behavior (response times should be similar across regions if CDN is working)
Interpreting location results
| Pattern | Likely cause |
|---|---|
| All locations: slow | Server is overloaded or hosting is slow globally |
| One region: slow | CDN node issue in that region, or no CDN configured for that region |
| One region: down | Regional routing issue, or the site is geo-blocked in that region |
| All locations: down | Server or hosting is down globally |
| Intermittent failures across regions | Server-side errors under load (race conditions, memory pressure) |
Multi-location vs. dual confirmation
These are two different features that both reduce false positives:
| Feature | How it works | Best for |
|---|---|---|
| Dual confirmation | Requires 2 consecutive failed checks from the primary server | Filtering momentary network blips (seconds) |
| Multi-location check | Verifies the failure from 6 different global locations | Confirming real vs. regional outages |
You can use both together. With dual confirmation + multi-location, an alert is only sent if the site fails two checks in a row and that failure is confirmed from multiple geographic regions.