On Saturday, August 05, 2017 21:37:35 Ultima wrote:
> Never tested SLAAC on lagg, but your configuration looks correct to me. One
> flag that I notice that looks questionable is the MTU option being 9216
> while the lagg interface is set to 1500. I doubt this would cause SLAAC to
> fail but it maybe worth investigating.
I had noticed that too. I'll ask the networking guys why the RA has that MTU 
option. As you say, I it's probably not the cause: the Linux hosts have have 
the same hardware and an analogous configuration using the bonding driver. The 
Linux bond interfaces also have MTU 1500, yet they SLAAC fine:

bond0     Link encap:Ethernet  HWaddr D0:BF:9C:F1:C9:E8  
          inet addr:xxx.xxx.212.13  Bcast:xxx.xxx.212.127  Mask:
          inet6 addr: 26xx:xxxx:4013:4:d2bf:9cff:fef1:c9e8/64 Scope:Global
          inet6 addr: fe80::d2bf:9cff:fef1:c9e8/64 Scope:Link
          RX packets:139503168 errors:0 dropped:0 overruns:0 frame:0
          TX packets:166569190 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:17165085983 (15.9 GiB)  TX bytes:90980707420 (84.7 GiB)

It may not be lagg related. As I recall, a few months ago I tried bringing one 
of the hosts up on oce0 alone, and SLAAC failed then too. I wonder if it's an 
oce(4) problem?

Greg Rivers
freebsd-stable@freebsd.org mailing list
To unsubscribe, send any mail to "freebsd-stable-unsubscr...@freebsd.org"

Reply via email to