Monitoring

Website Monitoring and Uptime Fundamentals

What to monitor, how often, from where, and what to alert on — covering uptime, performance, response time, SSL expiry, DNS drift, and the operational discipline that turns monitoring data into reliability outcomes.

Start monitoring sites

Quick summary

Poll every 60s from multiple regions; alert on 2+ failures across regions.
Monitor uptime AND content — a 200 with broken HTML is still an outage.
Track SSL expiry and DNS drift on slower cadences with separate alerts.
Status pages turn private monitoring into customer trust.

The four monitoring layers

Reachability — TCP/HTTP responds at all.
Correctness — expected status code and keyword on the page.
Performance — response time within SLO.
Configuration drift — SSL expiry, DNS changes, header regressions.

Alerting without burning out the team

Multi-region consensus before alerting. Group alerts by service, not by check. Suppress during planned maintenance windows. Route by severity — not every check is page-worthy.

Status pages and external trust

A public status page that updates from the same data the team uses converts incidents from a support storm into a confidence event. Honesty during outages buys more trust than uptime numbers ever do.

Troubleshooting workflow

Run through these steps in order. Each step links to the right diagnostic tool.

1
Define what 'up' means
Status code + keyword + latency budget.
Open Validate a site
2
Monitor SSL expiry
Alert 30/14/7 days out.
Open SSL Checker
3
Track DNS drift
Watch authoritative records over time.
Open DNS Lookup
4
Watch headers
Detect HSTS/CSP regressions.
Open Headers

Recommended next steps

Website downtime monitoring guide

Business uptime monitoring

SSL expiration monitoring

DNS monitoring guide

Network troubleshooting

Internet intelligence playbook

Frequently asked questions

What is the difference between uptime and availability?

Uptime is whether the service responds. Availability adds correctness — a 200 OK serving a broken page is uptime but not availability. Monitor both with keyword checks plus status checks.

How often should I poll?

60 seconds is the practical sweet spot for production. Faster wastes budget; slower lets short outages slip past.

Where should checks run from?

Multiple regions. A single-region monitor confuses a local network blip with a real outage and alerts on noise.

What should I alert on?

Two consecutive failures from two independent regions. Single-region single-failure alerts cause pager fatigue and erode trust in the system.

Should I monitor SSL and DNS too?

Yes. Expiring SSL and silent DNS drift cause outages that uptime checks notice late. Run them on a slower cadence with their own alerts.

Monitor it continuously

Save this check, get alerts on changes, and track multiple domains in one dashboard.

Learn about monitoring

Platforms that pair with this workflow

Curated tools we trust. Some links are sponsored — see our advertising disclosure.