IPv6 Link-Local Ping Fails Without Zone ID
Advanced TroubleshootQuestion
Two devices are connected on the same Ethernet link. Device A has link-local address fe80::1 on interface eth0. Device B has link-local address fe80::2 on interface eth1. Running 'ping6 fe80::2' from Device A fails. The devices are physically on the same link. Identify the problem and the correct ping command.
Given
{
"device_a": "fe80::1%eth0",
"device_b": "fe80::2%eth1",
"same_link": true
}
Answer
{
"problem": "Link-local addresses require a zone ID (scope identifier) to specify which interface to use for sending",
"wrong_command": "ping6 fe80::2",
"correct_command": "ping6 fe80::2%eth0",
"explanation": "Without %eth0, the OS does not know which interface to use to send the ping. Link-local addresses are not globally unique \u2014 every host has fe80::something on every interface.",
"fix": "Always append %<interface> when pinging IPv6 link-local addresses"
}
Try to solve it first, then click Reveal Answer
Step-by-Step Explanation
Step 1: Understand IPv6 link-local addresses
IPv6 link-local addresses (fe80::/10) are: - Automatically configured on every IPv6-capable interface - Only valid on a single link (not routed) - Not globally unique — multiple interfaces can have addresses in the same fe80::/64 range
Because they are link-scoped, the OS must know which interface to use when sending a packet to a link-local destination.
Step 2: Why 'ping6 fe80::2' fails
When you run ping6 fe80::2 without specifying an interface:
- OS looks up fe80::2 in the routing table
- No route for fe80::/10 exists in the unicast routing table (link-local routes are per-interface)
- OS cannot determine which interface to send the ICMP6 echo request on
- Error: "Network is unreachable" or "Cannot assign requested address"
If the host has multiple interfaces, the OS might pick the wrong one or fail entirely.
Step 3: The zone ID (scope identifier) syntax
RFC 4007 defines the zone ID concept. In practice:
fe80::2%eth0 # Linux — use eth0 interface
fe80::2%en0 # macOS — use en0 interface
fe80::2%4 # Windows — use interface index 4
The %<iface> suffix tells the OS which interface to use for sending the packet.
Step 4: Correct ping commands
From Device A (interface eth0):
# Linux / macOS
ping6 fe80::2%eth0
# Windows PowerShell
ping -6 fe80::2%4
# (replace 4 with actual interface index from 'netsh interface ipv6 show interfaces')
# Alternative: specify source interface
ping6 -I eth0 fe80::2
Step 5: How to find the link-local address and interface
On Device B:
# Linux
ip addr show | grep fe80
# → inet6 fe80::2/64 scope link
# macOS
ifconfig | grep fe80
# Windows
netsh interface ipv6 show addresses
Step 6: Verify connectivity
# From Device A — should now succeed
ping6 fe80::2%eth0
# PING fe80::2%eth0(fe80::2) 56 data bytes
# 64 bytes from fe80::2%eth0: icmp_seq=1 ttl=64 time=0.085 ms
Key lesson
Always use the %interface zone identifier with fe80:: addresses. Without it, the command is ambiguous on hosts with multiple interfaces. This is a frequent stumbling block for engineers transitioning from IPv4 to IPv6.
The zone ID is also required in URLs: http://[fe80::2%25eth0]/ (where %25 is URL-encoded %).