Description

v6 link-local address not working on EVO after 22.1

Symptoms

The issue was found on PTX10001-36MR, but it is common for all EVOs.

Testbed in the lab:

> show configuration interfaces et-0/0/10

flexible-vlan-tagging;

encapsulation flexible-ethernet-services;

unit 3777 {

vlan-id 3777;

family inet6 {

address fe80::bbbb/64;

}

}

unit 3888 {

vlan-id 3888;

family inet6 {

address fe80:d:e:a:d::bbbb/64;

}

}

unit 3999 {

vlan-id 3999;

family inet6 {

address fe80:b:e:e:f::bbbb/64;

}

}

 

 

> ping inet6 interface et-0/0/10.3777 fe80::bbbb count 3

/bin/ping: Warning: source address might be selected on device other than: et-0/0/10.3777

PING fe80::bbbb(fe80::bbbb) from :: et-0/0/10.3777: 56 data bytes

64 bytes from fe80::bbbb%vrf0: icmp_seq=1 ttl=64 time=4.69 ms

64 bytes from fe80::bbbb%vrf0: icmp_seq=2 ttl=64 time=0.334 ms

64 bytes from fe80::bbbb%vrf0: icmp_seq=3 ttl=64 time=0.611 ms

 

--- fe80::bbbb ping statistics ---

3 packets transmitted, 3 received, 0% packet loss, time 2041ms

rtt min/avg/max/mdev = 0.334/1.878/4.690/1.991 ms

 

> ping inet6 interface et-0/0/10.3888 fe80:d:e:a:d::bbbb count 3

/bin/ping: Warning: source address might be selected on device other than: et-0/0/10.3888

PING fe80:d:e:a:d::bbbb(fe80:d:e:a:d::bbbb) from :: et-0/0/10.3888: 56 data bytes

/bin/ping: sendmsg: Network is unreachable

/bin/ping: sendmsg: Network is unreachable

/bin/ping: sendmsg: Network is unreachable

^C

--- fe80:d:e:a:d::bbbb ping statistics ---

3 packets transmitted, 0 received, 100% packet loss, time 2078ms

 

 

> ping inet6 interface et-0/0/10.3999 fe80:b:e:e:f::bbbb count 3

/bin/ping: Warning: source address might be selected on device other than: et-0/0/10.3999

PING fe80:b:e:e:f::bbbb(fe80:b:e:e:f::bbbb) from :: et-0/0/10.3999: 56 data bytes

/bin/ping: sendmsg: Network is unreachable

/bin/ping: sendmsg: Network is unreachable

/bin/ping: sendmsg: Network is unreachable

 

^C

--- fe80:b:e:e:f::bbbb ping statistics ---

3 packets transmitted, 0 received, 100% packet loss, time 2026ms

 

 

The basic issue seems to be the link-local address used is not compliant with RFC. It should have network field fe80 followed by zero. (i.e. fe80::/64)

 

https://datatracker.ietf.org/doc/html/rfc4291#section-2.5.6

 

It works when I change the address from fe80:d:e:a:d::bbbb to fe80::e:a:d:bbbb so that the upper 64 bits are fe80::

 

There is no check if an invalid IPv6 link-local address is configured.

 

Solution

Need to delete the invalid ipv6 link-local address and configure a valid one

This is a limitation of the software. The limitation is fixed in later version.

Modification History

2024-01-31 : Article Created