During a recent domain migration, something strange started happening.
Some regions could reach the site perfectly. Others failed completely. And a few flipped between success and failure with no clear pattern.
At first glance, it looked like instability. But it wasn’t.
It was DNS.
The Problem: Inconsistent Failures Across Regions
We were monitoring a domain during a migration to a new host. Our system runs health checks every minute from multiple global regions, so we immediately started seeing conflicting results:
- North America → mostly successful
- Europe → intermittent failures
- Asia-Pacific → mixed behavior
Some checks returned:
- “DNS name not found”
- “connection timeout”
Others returned normal responses with healthy latency.
This is exactly the kind of situation that’s difficult to diagnose quickly. Traditional uptime monitoring would just show partial outages or noisy alerts, without much explanation.
What Most Monitoring Tools Would Show
In a typical setup, you’d see:
- A drop in uptime percentage
- A series of alerts from failing regions
- Maybe some raw error messages
But you’d still be left asking:
- Is this downtime?
- Is the server failing?
- Is it networking?
- Or something else entirely?
How Uptivus Looked at the Problem Differently
Uptivus doesn’t just look at whether a check passed or failed. It looks at patterns across regions over time.
Every minute, serverless functions deployed globally run checks and report back:
- Did the request succeed?
- How long did it take?
- What error occurred (if any)?
This builds a continuous stream of region-tagged data.
When something unusual happens, the system doesn’t immediately jump to conclusions. It first gathers more context.
The AI Insight That Made the Difference
Instead of treating each failure as an isolated event, Uptivus grouped the results by region and analyzed them collectively.
Here’s what it found:
- Certain regions had a high rate of DNS-related errors
- Other regions had zero issues
- The failures were not random — they were geographically clustered
From that pattern, the AI identified:
This is not a full outage. This is a regional DNS resolution issue.
In other words, the domain was resolving correctly in some parts of the world, but not others — a classic sign of DNS propagation during a migration.
Why This Matters
DNS issues during migrations are notoriously confusing because they don’t fail consistently.
You might see:
- Some users reporting downtime
- Others saying everything works fine
- Monitoring tools giving mixed signals
Without context, this leads to:
- Unnecessary panic
- Wasted debugging time
- False assumptions
By identifying the issue as regional DNS propagation, Uptivus turned a confusing situation into a clear explanation.
Monitoring isn’t just about detecting failures anymore.
It’s about understanding them.
The goal with Uptivus is simple:
- Not just to tell you that something is wrong
- But to tell you what is wrong and why