I have had some success in identifying the root cause of this. Today, IPv6 broke again, and I had some spare time to debug it.
It seems as though NetworkManager is setting up an IPv6 default route via em1, even though I had em1 disabled.
It seems as though NetworkManager is setting up an IPv6 default route via em1, even though I had em1 disabled.
Earlier, I had two default routes. One via em1, another via wlp2s0 (aka wlan0):
default via fe80::d267:e5ff:feb5:235b dev wlp2s0 proto static metric 1024
default via fe80::d267:e5ff:feb5:235b dev em1 proto ra metric 1024 expires 593sec
default via fe80::d267:e5ff:feb5:235b dev em1 proto ra metric 1024 expires 593sec
One appears to be static, the other is received via ra on the disabled interface.
I can only imagine this means my default route in IPv6 land has been switching between the two interfaces. To me, this explains the dropped ssh sessions, web browsing timeouts, and other flaky behaviour which I am attributing to loss of IPv6 DNS packages.
I can only imagine this means my default route in IPv6 land has been switching between the two interfaces. To me, this explains the dropped ssh sessions, web browsing timeouts, and other flaky behaviour which I am attributing to loss of IPv6 DNS packages.
So, NetworkManager bug or kernel bug?
I can only assume NetworkManager is responsible for this.
Cheers
I can only assume NetworkManager is responsible for this.
Cheers
On Thu, Jan 23, 2014 at 1:29 PM, Dan Irwin <rummymobile@xxxxxxxxx> wrote:
ThanksHi Wolfgang,I'm not using dhclient6, or dhcpv6 at all on the subnets I connect to. Checking my system I don't see dhclient6 running. Sounds like a similar problem though.
I might try the updated policy package mentioned in your bz - but only if the changelog seems like it might help.On Thu, Jan 23, 2014 at 10:07 AM, Wolfgang S. Rupprecht <wolfgang.rupprecht@xxxxxxxxx> wrote:
I'm seeing an IPv6 breakage in NetworkManager/dhclient6 too. At first
Dan Irwin <rummymobile@xxxxxxxxx> writes:
> Seeing significant issues with IPv6. After some time, IPv6 completely
> seems to stop working. IPv4 is unaffected.
it looked like an Selinux problem (there was an avc), but fixing that
didn't stop the problem with NM not getting a working IPv6 address via
dhclient6. BZ#1055226 https://bugzilla.redhat.com/show_bug.cgi?id=1055226
Mine is via an ethernet connection to my cable modem which connects to
> Re-connecting to wifi seems to restore IPv6 connectivity.
Comcast. I can do a "systemctl restart NetworkManager" in order to get
an ipv6 address.
-wolfgang
--
users mailing list
users@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Have a question? Ask away: http://ask.fedoraproject.org
-- users mailing list users@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe or change subscription options: https://admin.fedoraproject.org/mailman/listinfo/users Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines Have a question? Ask away: http://ask.fedoraproject.org
