Re: [lxc-users] "The configuration file contains legacy configuration keys" - which ones?

[Stéphane Graber]

On Wed, Sep 27, 2017 at 10:07:13PM +0900, Tomasz Chmielewski wrote:
> On 2017-09-27 22:03, Stéphane Graber wrote:
> > On Wed, Sep 27, 2017 at 09:48:39PM +0900, Tomasz Chmielewski wrote:
> > > # lxc exec some-container /bin/bash
> > > The configuration file contains legacy configuration keys.
> > > Please update your configuration file!
> > > 
> > > 
> > > 
> > > Is there a way to tell find out which ones are legacy without
> > > pasting the
> > > whole config on the mailing list?
> > > 
> > > 
> > > Tomasz Chmielewski
> > > https://lxadm.com
> > 
> > In most cases, it's just that the container was started on LXC 2.0.x, so
> > just restarting it will have LXD generate a new LXC 2.1 config for it,
> > getting you rid of the warning.
> > 
> > If that warning remains, then it means you're using a "raw.lxc" config
> > in your container or one of its profile and that this key is the one
> > which contains a now legacy config key.
> > 
> > 
> > Details on key changes can be found here:
> > 
> >   https://discuss.linuxcontainers.org/t/lxc-2-1-has-been-released/487
> > 
> > 
> > The tl is that for 99% of LXD users, all you need to do is restart
> > your running containers so that they get switched to the new config
> > format, no config change required. The remaining 1% is where you're
> > using raw.lxc which then needs manual updating to get rid of the
> > warning.
> 
> I use this one, as I have a newer kernel from Ubuntu ppa:
> 
> config:
>raw.lxc: lxc.aa_allow_incomplete=1
> 
> 
> So how exactly do I modify it?
> 
> lxc.aa_allow_incomplete   ->   lxc.apparmor.allow_incomplete

Yeah, edit your container (lxc config edit) or profile (lxc profile
edit) and just replace lxc.aa_allow_incomplete=1 with
lxc.apparmor.allow_incomplete=1 and you should be good to go.

> 
> ?
> 
> 
> Tomasz Chmielewski
> https://lxadm.com

-- 
Stéphane Graber
Ubuntu developer
http://www.ubuntu.com


signature.asc
Description: PGP signature
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users

Re: [lxc-users] "The configuration file contains legacy configuration keys" - which ones?

[Tomasz Chmielewski]

On 2017-09-27 22:03, Stéphane Graber wrote:

On Wed, Sep 27, 2017 at 09:48:39PM +0900, Tomasz Chmielewski wrote:

# lxc exec some-container /bin/bash
The configuration file contains legacy configuration keys.
Please update your configuration file!



Is there a way to tell find out which ones are legacy without pasting 
the

whole config on the mailing list?


Tomasz Chmielewski
https://lxadm.com


In most cases, it's just that the container was started on LXC 2.0.x, 
so

just restarting it will have LXD generate a new LXC 2.1 config for it,
getting you rid of the warning.

If that warning remains, then it means you're using a "raw.lxc" config
in your container or one of its profile and that this key is the one
which contains a now legacy config key.


Details on key changes can be found here:

  https://discuss.linuxcontainers.org/t/lxc-2-1-has-been-released/487


The tl is that for 99% of LXD users, all you need to do is restart
your running containers so that they get switched to the new config
format, no config change required. The remaining 1% is where you're
using raw.lxc which then needs manual updating to get rid of the
warning.


I use this one, as I have a newer kernel from Ubuntu ppa:

config:
   raw.lxc: lxc.aa_allow_incomplete=1


So how exactly do I modify it?

lxc.aa_allow_incomplete   ->   lxc.apparmor.allow_incomplete

?


Tomasz Chmielewski
https://lxadm.com
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users

Re: [lxc-users] "The configuration file contains legacy configuration keys" - which ones?

[Stéphane Graber]

On Wed, Sep 27, 2017 at 09:48:39PM +0900, Tomasz Chmielewski wrote:
> # lxc exec some-container /bin/bash
> The configuration file contains legacy configuration keys.
> Please update your configuration file!
> 
> 
> 
> Is there a way to tell find out which ones are legacy without pasting the
> whole config on the mailing list?
> 
> 
> Tomasz Chmielewski
> https://lxadm.com

In most cases, it's just that the container was started on LXC 2.0.x, so
just restarting it will have LXD generate a new LXC 2.1 config for it,
getting you rid of the warning.

If that warning remains, then it means you're using a "raw.lxc" config
in your container or one of its profile and that this key is the one
which contains a now legacy config key.


Details on key changes can be found here:

  https://discuss.linuxcontainers.org/t/lxc-2-1-has-been-released/487


The tl is that for 99% of LXD users, all you need to do is restart
your running containers so that they get switched to the new config
format, no config change required. The remaining 1% is where you're
using raw.lxc which then needs manual updating to get rid of the
warning.

-- 
Stéphane Graber
Ubuntu developer
http://www.ubuntu.com


signature.asc
Description: PGP signature
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users

Re: [lxc-users] "The configuration file contains legacy configuration keys" - which ones?

[Saint Michael]

I add myself to the question.

On Wed, Sep 27, 2017 at 8:48 AM, Tomasz Chmielewski  wrote:

> # lxc exec some-container /bin/bash
> The configuration file contains legacy configuration keys.
> Please update your configuration file!
>
>
>
> Is there a way to tell find out which ones are legacy without pasting the
> whole config on the mailing list?
>
>
> Tomasz Chmielewski
> https://lxadm.com
> ___
> lxc-users mailing list
> lxc-users@lists.linuxcontainers.org
> http://lists.linuxcontainers.org/listinfo/lxc-users
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users

[lxc-users] "The configuration file contains legacy configuration keys" - which ones?

[Tomasz Chmielewski]

# lxc exec some-container /bin/bash
The configuration file contains legacy configuration keys.
Please update your configuration file!



Is there a way to tell find out which ones are legacy without pasting 
the whole config on the mailing list?



Tomasz Chmielewski
https://lxadm.com
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users

Re: [lxc-users] Steam BPM X360 controller detection inside container

[Michael Honeyman]

Thanks for your reply Stéphane.

Using udevadm monitor I observed KERNEL events propagating inside the
container when the controller was attached (I imagined this was because the
container is running in privileged mode), so I thought there might be a way
to manually create a uevent inside the container using these.

Grateful for your advice, I will return the GUI functions of my system to
the host level :)

Michael.


> -- Forwarded message --
> From: "Stéphane Graber" 
> To: LXC users mailing-list 
> Cc:
> Bcc:
> Date: Tue, 26 Sep 2017 09:06:45 -0400
> Subject: Re: [lxc-users] Steam BPM X360 controller detection inside
> container
> On Sat, Sep 23, 2017 at 01:04:41PM +1000, Michael Honeyman wrote:
> > Hi lxc-users,
> >
> > I'm hoping someone wise can help me with some ideas how to solve this
> > problem.
> >
> > I've been having fun trying to get Steam in Big Picture Mode working
> inside
> > a LXD privileged container. The host is Ubuntu 16.04, as is the guest. I
> > have almost everything working (albeit slightly manually) inside the LXD
> > container including audio, video, and X360 wireless controller
> pass-through.
> >
> > My final challenge is with the addition of X360 wireless controllers
> during
> > execution of Steam. It will pick up controllers that were activated prior
> > to starting Steam, but once inside Steam BPM, a new controller or a
> > reconnected controller will not work until Steam is closed and
> re-launched.
> > If I start Steam, connect a controller (which then doesn't work in the
> > Steam interface), and then launch a game - the controller works inside
> the
> > game. So each time a new Steam executable is launched it will pick up the
> > newly attached controllers for that instance.
> >
> > I'm searching for some advice on what mechanism(s) Steam might be using
> to
> > detect newly added controllers, or how I might find out what it uses. I'm
> > hoping that if I can find out what Steam is expecting upon addition of a
> > new controller I can hopefully manually trigger that event inside the
> > container. My uninformed guess is Steam hooks into udev, but since udev
> > does not run inside containers this event is getting missed. But I'm not
> > really sure how to go about testing this theory, or any other methods it
> > might use.
> >
> > I realise this setup is very much on the fringe of the LXD use-case, so
> I'm
> > not hoping to find someone who has already solved this problem, just any
> > tips people can share to help me find the answer myself.
> >
> > Thanks in advance for any help and guidance rendered!
> >
> > Michael.
>
> My guess is that steam is looking for a hotplug uevent for the USB
> device, using this to detect when a device is plugged in after startup.
>
> Unfortunately containers don't receive uevents when a device is
> hotplugged, so if that's what steam's doing, you're out of luck for now.
>
> --
> Stéphane Graber
> Ubuntu developer
> http://www.ubuntu.com
>
>
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users

Re: [lxc-users] containers and lxdbr0 bridge lost ip after upgrade on 2.18

[Laurent Ducos]

Yes it's a dnsmasq version parsing issue
Thanks for the fix
26 septembre 2017 15:07 "Stéphane Graber"  a écrit:
> On Fri, Sep 22, 2017 at 02:29:15PM +, Laurent Ducos wrote:
> 
>> Hello 
>> After xenial ubuntu upgrade lxd switch to  2.18
>> i reboot and no ip for lxdbr0 and no ip for container
>> It set manualy ip for lxdbr0 "sudo ip addr add dev lxdbr0 10.215.181.1/24" 
>> and launch a container
>> but  always no ip 
>> Here is syslog entry when i try to start the container
>> 
>> Sep 22 16:18:07 blabla kernel: [ 3482.496088] lxdbr0: port 1(vethFVYDNI) 
>> entered disabled state
>> Sep 22 16:18:07 blabla kernel: [ 3482.501486] device vethFVYDNI left 
>> promiscuous mode
>> Sep 22 16:18:07 blabla kernel: [ 3482.501490] lxdbr0: port 1(vethFVYDNI) 
>> entered disabled state
>> Sep 22 16:18:08 blabla kernel: [ 3483.329216] audit: type=1400 
>> audit(1506089888.350:166):
>> apparmor="STATUS" operation="profile_remove" profile="unconfined" 
>> name="lxd-user10-blabla-sandbox_"
>> pid=77593 comm="apparmor_parser"
>> Sep 22 16:18:16 blabla kernel: [ 3491.046769] audit: type=1400 
>> audit(1506089896.070:167):
>> apparmor="STATUS" operation="profile_load" profile="unconfined" 
>> name="lxd-user10-blabla-sandbox_"
>> pid=77629 comm="apparmor_parser"
>> Sep 22 16:18:16 blabla systemd-udevd[77631]: Could not generate persistent 
>> MAC address for
>> vethB22DLS: No such file or directory
>> Sep 22 16:18:16 blabla kernel: [ 3491.057615] device vethCORADG entered 
>> promiscuous mode
>> Sep 22 16:18:16 blabla kernel: [ 3491.057808] IPv6: ADDRCONF(NETDEV_UP): 
>> vethCORADG: link is not
>> ready
>> Sep 22 16:18:16 blabla kernel: [ 3491.146758] eth0: renamed from vethB22DLS
>> Sep 22 16:18:16 blabla kernel: [ 3491.167857] IPv6: ADDRCONF(NETDEV_CHANGE): 
>> vethCORADG: link
>> becomes ready
>> Sep 22 16:18:16 blabla kernel: [ 3491.167961] lxdbr0: port 1(vethCORADG) 
>> entered forwarding state
>> Sep 22 16:18:16 blabla kernel: [ 3491.167971] lxdbr0: port 1(vethCORADG) 
>> entered forwarding state
>> Sep 22 16:18:31 blabla kernel: [ 3506.214893] lxdbr0: port 1(vethCORADG) 
>> entered forwarding state
>> my network configuration
>> config:
>> dns.domain: blabla.local
>> dns.mode: dynamic
>> ipv4.address: 10.215.181.1/24
>> ipv4.nat: "true"
>> description: ""
>> name: lxdbr0
>> type: bridge
>> used_by:
>> - /1.0/containers/lxd-user10-blabla-sandbox
> 
> Hi,
> 
> That sounds like the dnsmasq version parsing issue we fixed a few days ago.
> 
> Can you paste "dnsmasq --version" so we can make sure the new code
> handles your output properly?
> 
> --
> Stéphane Graber
> Ubuntu developer
> http://www.ubuntu.com
> 
> ___
> lxc-users mailing list
> lxc-users@lists.linuxcontainers.org
> http://lists.linuxcontainers.org/listinfo/lxc-users
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
http://lists.linuxcontainers.org/listinfo/lxc-users