2025 Year in Review: The State of Internet Uptime

2025 was a chaotic year for the internet. We saw major outages from cloud providers, massive DDoS attacks, and the rise of AI agents causing accidental Denial of Service events.

At Cluster Uptime, we processed over 5 Billion uptime checks this year. Here is what the data tells us.

1. The “Big Three” Report Card

How did the major cloud providers fare?

• AWS: 99.98% Global Uptime.
  • Major Incident: The us-east-1 DynamoDB latency spike in March caused cascading failures for 4 hours.
• GCP: 99.99% Global Uptime.
  • Stability: Remarkably stable this year, winning the reliability crown.
• Azure: 99.95% Global Uptime.
  • Issues: Several regional DNS failures in Europe impacted uptime significantly.

2. The Rise of “Micro-Outages”

We noticed a trend: Outages are getting shorter but more frequent. Instead of a 4-hour blackout, we see 10-second “blips” happening 50 times a day.

• Cause: Automated redeployments (Kubernetes/CI/CD).
• Impact: Persistent connections (WebSockets) are dropped, frustrating users, even if the “Uptime” looks green.

3. IPv6 Adoption

For the first time, over 40% of our checks were performed over IPv6. If you aren’t monitoring your IPv6 endpoints specifically, you might be missing outages that only affect mobile users (who are often IPv6-only via NAT64).

4. Fastest DNS Providers

We measured the DNS Resolution Time for millions of checks.

1. Cloudflare (1.1.1.1): 12ms avg.
2. Google (8.8.8.8): 18ms avg.
3. ISPs (Comcast/AT&T): 45ms avg. (Still slow!).

5. Security & DDoS

DDoS attacks increased by 200% this year. Monitoring agents often detected the attack before the security tools did, simply by noticing a massive spike in “Time to Connect” (TCP Handshake latency) as the firewall struggled to process new connections.

Looking Ahead to 2026

The internet is getting faster but more fragile. Complexity is increasing. The only defense is Observability. You can’t fix what you can’t measure.

See you in 2026!