AWS, Google Cloud & Vercel uptime monitoring
What CloudWatch Synthetics, Route 53 health checks and Google Cloud uptime checks do well, where they fall short, and when an external monitor is worth adding.
· 5 min read · By the Spot Downtime team
Every big cloud has its own way to check that your app is up. They're good tools, and if you already live in one console they're the obvious first stop. But a check that runs inside the same provider as your app shares its blind spots. Here's what each one offers and when to add an outside view.
What the platforms give you
| Platform | Built-in option | Good at | Watch out for |
|---|---|---|---|
| AWS | Route 53 health checks, CloudWatch Synthetics canaries | Failover routing, scripted browser checks, alarms into CloudWatch | Priced per check or canary run; alerts go through CloudWatch and SNS, in AWS |
| Google Cloud | Cloud Monitoring uptime checks | Checks from several regions, alerting policies, notification channels | Set up per project; alerts go through Google Cloud |
| DigitalOcean | Uptime checks | Simple HTTP checks and alerts in the control panel | Covers what you host on DigitalOcean |
| Vercel, Netlify | Deploy checks, logs and analytics | Catching a bad deploy | Not a schedule of checks on the live site from outside |
Google Cloud's alerting can notify by email, SMS, Slack, PagerDuty, webhooks, Pub/Sub and the Cloud Console mobile app; you add these as notification channels on an alerting policy, then attach the policy to the uptime check.
Where a built-in check falls short
- Shared failure. When the provider has a bad day, the thing that tells you about it can be affected too: the console, the alarm pipeline, the notifications. An outside monitor keeps working.
- Only half your stack. Your app on AWS also depends on DNS at one company, a CDN at another, email at a third and payments at a fourth. One external tool can watch all of them, plus the provider's own status page.
- One check per account. Several projects, clients or clouds means several consoles. An external monitor puts them on one dashboard and one status page.
- Cost at small scale. Per-check pricing adds up across many endpoints and short intervals. Many external tools have a free tier.
A setup that uses both
- Keep Route 53 health checks where they drive DNS failover. That's their real job.
- Use CloudWatch or Cloud Monitoring for internal metrics: CPU, queue depth, error rates.
- Add an external uptime monitor for the public URLs your customers use, checked from several regions, with alerts that don't depend on the provider.
- Watch the provider's status for the regions you use, so you know straight away when a problem is theirs.
Doing the outside part with Spot Downtime
Add a Website or API monitor for each public URL, whichever platform it runs on. Then add Third-party service monitors for AWS, Google Cloud, Azure, DigitalOcean, Vercel or Netlify, picking just the regions and components you use, so a provider incident shows up next to your own checks. Both kinds are on the Free plan.
Checking a cloud provider right now?
Get new guides by email
Practical posts on monitoring, outages, error fixes and on-call, when they're published. No spam.
Keep reading
- Monitoring · ComparisonsWebsite monitoring tools: the six kinds, and which ones you needUptime, performance, errors, changes, security and traffic: what each kind of website monitoring software watches, what it misses, and the smallest set that covers a real site.November 3, 2026 · 5 min read
- Monitoring · ComparisonsOpen-source website monitoring: Zabbix, Grafana, Uptime Kuma and GatusHow to monitor websites with open-source tools: Zabbix web scenarios, Prometheus and Grafana with the blackbox exporter, Uptime Kuma and Gatus, with config examples and the catch every self-hosted monitor shares.November 3, 2026 · 5 min read