0

ip addr shows:

2:  eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
     link/ether 02:6d:5e:86:29:49 brd ff:ff:ff:ff:ff:ff
     inet X.X.X.X/24 brd X.X.X.255 scope global eth0
        valid_lft forever preferred_lft forever
     inet6 2604:X:X:X::11/64 scope global deprecated
        valid_lft forever preferred_lft 0sec
     inet6 fe80::X:X:X:X/64 scope link
        valid_lft forever preferred_lft forever

In particular, I'm wondering about the deprecated and preferred_lft 0sec in there.

My syslog has this repeated over and over:

dhclient: PRC: Address 2604:X:X:X::11 depreferred.
dhclient: PRC: Address 2604:X:X:X::11 expired.
dhclient: PRC: Bound lease is devoice of active addresses. Re-initializing.
dhclient: XMT: Solicit on eth0, interval 1010ms.
dhclient: RCV: Advertise message on eth0 from ::.
dhclient: XMT: Request on eth0, interval 920ms.
dhclient: RCV: Reply message on eth0 from ::.

Then about four seconds later, it does it again. I can see the address appearing in and disappearing from the ip addr output.

I can ping the gateway while the address is present:

64 bytes from 2604:X:X:X::1: icmp_seq=1 ttl=255 time=0.318 ms
64 bytes from 2604:X:X:X::1: icmp_seq=2 ttl=255 time=0.277 ms
64 bytes from 2604:X:X:X::1: icmp_seq=3 ttl=255 time=0.336 ms
64 bytes from 2604:X:X:X::1: icmp_seq=4 ttl=255 time=0.324 ms
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
64 bytes from 2604:X:X:X::1: icmp_seq=7 ttl=255 time=0.316 ms
64 bytes from 2604:X:X:X::1: icmp_seq=8 ttl=255 time=0.321 ms
64 bytes from 2604:X:X:X::1: icmp_seq=9 ttl=255 time=0.331 ms
64 bytes from 2604:X:X:X::1: icmp_seq=10 ttl=255 time=0.327 ms

But if I try and ping this host from the Internet, it sees

IP6 :: > ff02:1:ff00:11: ICMP6, neighbor solicitation, who has 2604:X:X:X::11, length 24

multiple times, but never sends a response. ip maddr show does list inet6 ff02::1:ff00:11 on eth0.

At this point I've run out of things to look for. What might the problem be? In particular, what's going on with that immediately-expiring address from DHCP (it is the correct address, btw)?

Edit: I looked at the tcpdump in Wireshark, and I see that the leases are being given with a preferred lifetime of 4s and a valid lifetime of 8s, which is consistent with the timing in the syslog.

This is actually a VM running on an OpenContrail network, but I want to rule out it simply being a Linux issue first.

Isvara
  • 215
  • 1
  • 12
  • 1
    The DHCPv6 server is misconfigured. Find whoever runs it, and make them fix it. – Michael Hampton Apr 30 '16 at 23:35
  • Is that causing the NDP problem too, though? I'd expect it to respond during the times when the address is on the interface, just like it responds to pings during those moments. – Isvara May 01 '16 at 00:40

0 Answers0