Public bug reported:
The metadata server listens on 169.254.269.254 but not on any ipv6
address. Thus v6-only machines must be configured with a config-drive.
(There is no v6 address assigned to metadata, but it could easily listen
on the v4-in-v6 address ::169.254.169.254 in the same way.)
**
Public bug reported:
The metadata server listens on 169.254.269.254 but not on any ipv6
address. Thus v6-only machines must be configured with a config-drive.
(There is no v6 address assigned to metadata, but it could easily listen
on the v4-in-v6 address ::169.254.169.254 in the same way.)
**
Seems to me that the issue is not that the MTU isn't being honoured, but
that the MTU should be checked *before* TSO assembly and not *after*.
Assembly should happen outside the VM if the VM has enabled it on the
interface internally, and clearly the incoming segments (not packets, at
this point)
Seems to me that the issue is not that the MTU isn't being honoured, but
that the MTU should be checked *before* TSO assembly and not *after*.
Assembly should happen outside the VM if the VM has enabled it on the
interface internally, and clearly the incoming segments (not packets, at
this point)
We sure this isn't the issue Aaron's reported whereby if a machine is
scheduled, attempts to start, fails to start and is rescheduled we can
sometimes create a second port? (To be solved, but last I heard not yet
solved, by creating the port before scheduling, rather than in the
compute node.)
It's a little more than wishlist, I would argue - I can find *no*
canonical, working documentation for the bonding options. bond-master
versus bond-slave, allow-bond0, etc. - nothing quite seems to work as
expected, because I'm guessing, because I'm googling and getting
documentation over a span
Additional data:
- You can't set get_gateway='none' as a kernel command line option; it's
documented as supported, but it doesn't work and you get an onscreen complaint
that the network configuration is not valid
- You can't netboot if your DHCP server doesn't supply a router option without
... looking at static.c it appears someone's added a reachability test
that's always run on the result of get_gateway(), even when the gateway
is explicitly not set.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
I'm debate the severity of the bug. If it's not about to be fixed (and
noting that, like Rudd-O, I've had this working properly, so it's not a
'feature' of SYS_AUTH) then it needs to be a lot clearer to people
installing NFS that mixed-userlist machines should not be expected to
work correctly.
Exploring the bug, it seems that rpc.idmap on the server maps a UID to a
name for RPC, and on the client does the reverse. The only earthly
reason for this is to get the server/client names straight.
Unfortunately it isn't applied on all operations - specifically, not
when you open a file - and
I was testing with devstack, and the latest Folsom of the moment. If
it's not turning up in either case at the moment then apparently it's
gone away - I'll reopen with more detail if I see it again.
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is
I was testing with devstack, and the latest Folsom of the moment. If
it's not turning up in either case at the moment then apparently it's
gone away - I'll reopen with more detail if I see it again.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is
My issue with this is that I think apt-add-repository should respect the
http proxy settings in /etc/apt/apt.conf (or should have its own proxy
settings if that's preferable), rather than explicitly requiring the
proxy to be supplied in the environment.
Also, for those that have issues with the
Public bug reported:
If I have the following in a libvirt.xml file:
serial type=unix
source path=/opt/stack/nova/serial/instance-0033-serial1
mode=bind/
target port=1/
/serial
Then I should end up with a unix socket created when I hand this XML off
to libvirt. (I use
14 matches
Mail list logo
