IPv4 vs IPv6 latency: why results can differ

If you ping the same server over IPv4 and IPv6 and get different times, nothing is necessarily broken.

Why the numbers differ

• Different routes. IPv4 and IPv6 traffic often take different paths because providers peer differently for each protocol.
• Tunnelling and translation. Some networks still wrap IPv6 inside IPv4 (or the reverse), which adds overhead.
• Maturity of the setup. Older IPv4 infrastructure is heavily tuned. A newer IPv6 deployment may be faster, slower or simply less optimised.

How to test properly

1. Ping the host with each protocol several times, not once.
2. Compare the average and the maximum, not a single reply.
3. Run a traceroute for both and compare the paths.

When it matters

Modern browsers prefer IPv6 when it works and quietly fall back to IPv4 when it does not. If IPv6 is much slower or flaky, users can see delays before the fallback kicks in. That is worth checking whenever you publish an AAAA record.

The practical rule: if your site has both A and AAAA records, test and monitor both.