ping column cannot differentiate IPv4 and IPv6 #121
Description
https://lowendspirit.com/discussion/comment/230355/#Comment_230355
Device Under Test (DUT) is a KVM server physically located in Amsterdam NL.
DUT is an IPv6-only server and does not have a native IPv4 address.
It has an ip6gre tunnel to a private dual-stack router located in Bern CH, which providers NAT IPv4 connectivity to DUT.
YABS produced the following report on DUT:
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
# Yet-Another-Bench-Script #
# v2025-04-20 #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
Sun Mar 15 15:40:27 UTC 2026
Basic System Information:
---------------------------------
Uptime : 0 days, 0 hours, 49 minutes
Processor : AMD EPYC 7702P 64-Core Processor
CPU cores : 2 @ 1996.250 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 7.7 GiB
Swap : 512.0 MiB
Disk : 199.8 GiB
Distro : Debian GNU/Linux 13 (trixie)
Kernel : 6.12.74+deb13+1-amd64
VM Type : KVM
IPv4/IPv6 : ✔ Online / ✔ Online
IPv6 Network Information:
---------------------------------
ISP : Hollander \u0026 Jacobsen UG (haftungsbeschraenkt)
ASN : AS206075 Hollander \u0026 Jacobsen UG (haftungsbeschraenkt)
Host : Hollander \u0026 Jacobsen UG
Location : Amsterdam, North Holland (NH)
Country : Netherlands
fio Disk Speed Tests (Mixed R/W 50/50) (Partition /dev/sda3):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 249.62 MB/s (62.4k) | 3.49 GB/s (54.6k)
Write | 250.28 MB/s (62.5k) | 3.51 GB/s (54.9k)
Total | 499.91 MB/s (124.9k) | 7.01 GB/s (109.5k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 6.97 GB/s (13.6k) | 3.00 GB/s (2.9k)
Write | 7.34 GB/s (14.3k) | 3.20 GB/s (3.1k)
Total | 14.31 GB/s (27.9k) | 6.21 GB/s (6.0k)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 934 Mbits/sec | 952 Mbits/sec | 6.22 ms
Eranium | Amsterdam, NL (100G) | 951 Mbits/sec | 967 Mbits/sec | 0.529 ms
Uztelecom | Tashkent, UZ (10G) | 719 Mbits/sec | 412 Mbits/sec | 88.3 ms
Leaseweb | Singapore, SG (10G) | 572 Mbits/sec | 627 Mbits/sec | --
Clouvider | Los Angeles, CA, US (10G) | 534 Mbits/sec | 398 Mbits/sec | 142 ms
Leaseweb | NYC, NY, US (10G) | 770 Mbits/sec | 629 Mbits/sec | 85.8 ms
Edgoo | Sao Paulo, BR (1G) | 529 Mbits/sec | 520 Mbits/sec | 185 ms
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Clouvider | London, UK (10G) | 1.04 Gbits/sec | 979 Mbits/sec | 6.23 ms
Eranium | Amsterdam, NL (100G) | 1.04 Gbits/sec | 983 Mbits/sec | 0.526 ms
Uztelecom | Tashkent, UZ (10G) | 889 Mbits/sec | busy | 88.4 ms
Leaseweb | Singapore, SG (10G) | 604 Mbits/sec | 879 Mbits/sec | --
Clouvider | Los Angeles, CA, US (10G) | 597 Mbits/sec | 893 Mbits/sec | 142 ms
Leaseweb | NYC, NY, US (10G) | 898 Mbits/sec | 860 Mbits/sec | --
Edgoo | Sao Paulo, BR (1G) | 664 Mbits/sec | 787 Mbits/sec | 185 ms
Geekbench 6 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 1222
Multi Core | 2140
Full Test | https://browser.geekbench.com/v6/cpu/17086213
YABS completed in 14 min 41 sec
Notably, the IPv4 ping and IPv6 ping to a target located in Amsterdam NL are both less than 1ms:
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Eranium | Amsterdam, NL (100G) | 951 Mbits/sec | 967 Mbits/sec | 0.529 ms
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed | Ping
----- | ----- | ---- | ---- | ----
Eranium | Amsterdam, NL (100G) | 1.04 Gbits/sec | 983 Mbits/sec | 0.526 ms
However, given the networking setup of DUT, this cannot be the truth.
Manually pinging from DUT to the same target over IPv4 and IPv6 gives the following results:
sunny@ixp5:~$ ping -4 -c4 iperf-ams-nl.eranium.net
PING iperf-ams-nl.eranium.net (217.18.95.142) 56(84) bytes of data.
64 bytes from ptr.eranium.net (217.18.95.142): icmp_seq=1 ttl=56 time=29.8 ms
64 bytes from ptr.eranium.net (217.18.95.142): icmp_seq=2 ttl=56 time=30.1 ms
64 bytes from ptr.eranium.net (217.18.95.142): icmp_seq=3 ttl=56 time=30.1 ms
64 bytes from ptr.eranium.net (217.18.95.142): icmp_seq=4 ttl=56 time=30.2 ms
--- iperf-ams-nl.eranium.net ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3004ms
rtt min/avg/max/mdev = 29.789/30.039/30.168/0.149 ms
sunny@ixp5:~$ ping -6 -c4 iperf-ams-nl.eranium.net
PING iperf-ams-nl.eranium.net (2a09:e244:7000::142) 56 data bytes
64 bytes from 2a09:e244:7000::142: icmp_seq=1 ttl=59 time=0.742 ms
64 bytes from 2a09:e244:7000::142: icmp_seq=2 ttl=59 time=0.626 ms
64 bytes from 2a09:e244:7000::142: icmp_seq=3 ttl=59 time=0.562 ms
64 bytes from 2a09:e244:7000::142: icmp_seq=4 ttl=59 time=0.595 ms
--- iperf-ams-nl.eranium.net ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3047ms
rtt min/avg/max/mdev = 0.562/0.631/0.742/0.067 ms
Given the round-trip between DUT and the private router in Bern CH is 13ms, the IPv4 ping of 30ms is more likely to be true than the <1ms reported by YABS.