Global Network Test

Test website and server connectivity, latency and response time from multiple locations worldwide.

Run a global network test

Automatic uses HTTP for a full URL and Ping for a hostname or IP. TCP checks one explicit port.

What the Global Network Test checks

Run the same HTTP, Ping, or TCP connectivity test from multiple independent probe locations. Results arrive as each Agent finishes, so regional failures and latency differences are visible without waiting for every location.

Understanding the results

HTTP tests separate DNS lookup, TCP connection, TLS negotiation, time to first byte, and total response time. Ping reports round-trip latency and packet loss. TCP measures whether one specified public service port accepts a connection.

Common causes of regional differences

Longer physical paths, ISP peering, CDN routing, DNS answers, firewalls, and regional service incidents can produce different results. A failed Ping does not prove that a website is offline because some servers block ICMP while still serving HTTP.

Temporary sharing

Every test receives a private-by-link result URL. The request, results, and share mapping expire automatically after 14 days and shared pages are excluded from search indexing.

Frequently asked questions

Where are the tests run from?

Each result comes from an enabled NetOpsKit Agent in the location shown. Only healthy Agents with available capacity are selected.

Which test should I choose?

Use HTTP for websites and APIs, Ping for ICMP reachability and latency, or TCP to check whether one specific service port accepts a connection.

Why do locations return different results?

Physical distance, routing, peering, regional filtering, CDN selection, and temporary congestion can all change latency and availability.

How long is a shared result available?

Probe requests and share links are stored temporarily for 14 days, then Redis removes them automatically.

Can this tool test private servers?

No. Private, loopback, link-local, multicast, reserved, and cloud metadata targets are blocked before dispatch and checked again by every Agent.