Re: [Lxc-users] nilfs

2012-03-02 Thread Daniel Baumann
On 03/02/2012 08:45 AM, Ulli Horlacher wrote:
 I have tested btrfs on a standard Ubuntu 10.04.3 and one with kernel
 2.6.38-13-server (backport).

i'm not claiming btrfs is there yet, however, if you're using btrfs, you
should at least make sure to use something remotely up2date, say 3.2.x.
no surprise that btrfs in a one year old kernel is crap.

-- 
Address:Daniel Baumann, Donnerbuehlweg 3, CH-3012 Bern
Email:  daniel.baum...@progress-technologies.net
Internet:   http://people.progress-technologies.net/~daniel.baumann/

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] nilfs

2012-03-02 Thread Ulli Horlacher
On Fri 2012-03-02 (09:02), Daniel Baumann wrote:

 i'm not claiming btrfs is there yet, however, if you're using btrfs, you
 should at least make sure to use something remotely up2date, say 3.2.x.

SLES11 SP2 was released this week with a 3.0 kernel and comes with btrfs.
Same b(*CENSORED*)t as always from SuSE. What they label as Enterprise
is Testing on Debian.



-- 
Ullrich Horlacher  Server- und Arbeitsplatzsysteme
Rechenzentrum  E-Mail: horlac...@rus.uni-stuttgart.de
Universitaet Stuttgart Tel:++49-711-685-65868
Allmandring 30 Fax:++49-711-682357
70550 Stuttgart (Germany)  WWW:http://www.rus.uni-stuttgart.de/
REF: 4f507e79.1030...@progress-technologies.net

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


[Lxc-users] what is the best way to make dvb devices available in a container

2012-03-02 Thread Dieter Bloms
Hi,

I'am experimenting with lxc and have successfully make the dvb devices from
/dev/dvb/adapter0/* available to a container with the mount --bind
option.

But when I unload and load the drivers, the /dev/dvb of the
container is empty, because during unloading the drivers the /dev/dvb
directory of the host is removed and then during loading the drivers a
new one is created.

Is there a better way to make the dvb devices available in a container,
so the devices will exist in the container as they were created in the
host ?


-- 
best regards

  Dieter

--
I do not get viruses because I do not use MS software.
If you use Outlook then please do not put my email address in your
address-book so that WHEN you get a virus it won't use my address in the
From field.

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] adding a default gateway inside a container as a non root user

2012-03-02 Thread Daniel Lezcano
On 03/02/2012 11:18 AM, Arun M wrote:


 With the 0.8.0 version, you will be able to setup the gateway directly
 from the configuration file.

 Cool. I wanted to try this so tested with the latest code from repository.

 However hit another issue now.

 $ lxc-execute -n alpha -f n1.conf -l DEBUG -o log -- /bin/busybox ash
 lxc-execute: No such file or directory - failed to rename cgroup 
 /cgroup//lxc/3784-/cgroup//lxc/alpha
 ...

 lxc-execute 1331137335.969 INFO lxc_cgroup - [1] found cgroup mounted at 
 '/cgroup',opts='rw,relatime,blkio,net_cls,freezer,devices,memory,cpuacct,cpu,ns,cpuset'
 lxc-execute 1331137335.969 DEBUGlxc_cgroup - get_init_cgroup: found init 
 cgroup for subsys (null) at /
 lxc-execute 1331137335.969 DEBUGlxc_cgroup - cgroup /cgroup has flags 0x1
 lxc-execute 1331137335.969 WARN lxc_cgroup - using deprecated ns_cgroup
 lxc-execute 1331137335.969 ERRORlxc_cgroup - No such file or directory - 
 failed to rename cgroup /cgroup//lxc/3840-/cgroup//lxc/alpha

 $ file /cgroup/3840
 /cgroup/3840: directory

 $ file /cgroup/lxc/3840
 /cgroup/lxc/3840: cannot open `/cgroup/lxc/3840' (No such file or directory)

 It appears that lxc-execute is trying to look for a dir under /cgroup/lxc 
 while its actually present under /cgroup.


 Could not try cgroup.clone_children since it looks like this is not available 
 in kernel 2.6.32.

Ok, I see. I will look at fixing this.

   -- Daniel


--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] adding a default gateway inside a container as a non root user

2012-03-02 Thread Daniel Lezcano
On 03/02/2012 11:18 AM, Arun M wrote:


 With the 0.8.0 version, you will be able to setup the gateway directly
 from the configuration file.

 Cool. I wanted to try this so tested with the latest code from repository.

 However hit another issue now.

 $ lxc-execute -n alpha -f n1.conf -l DEBUG -o log -- /bin/busybox ash
 lxc-execute: No such file or directory - failed to rename cgroup 
 /cgroup//lxc/3784-/cgroup//lxc/alpha
 ...

 lxc-execute 1331137335.969 INFO lxc_cgroup - [1] found cgroup mounted at 
 '/cgroup',opts='rw,relatime,blkio,net_cls,freezer,devices,memory,cpuacct,cpu,ns,cpuset'
 lxc-execute 1331137335.969 DEBUGlxc_cgroup - get_init_cgroup: found init 
 cgroup for subsys (null) at /
 lxc-execute 1331137335.969 DEBUGlxc_cgroup - cgroup /cgroup has flags 0x1
 lxc-execute 1331137335.969 WARN lxc_cgroup - using deprecated ns_cgroup
 lxc-execute 1331137335.969 ERRORlxc_cgroup - No such file or directory - 
 failed to rename cgroup /cgroup//lxc/3840-/cgroup//lxc/alpha

 $ file /cgroup/3840
 /cgroup/3840: directory

 $ file /cgroup/lxc/3840
 /cgroup/lxc/3840: cannot open `/cgroup/lxc/3840' (No such file or directory)

 It appears that lxc-execute is trying to look for a dir under /cgroup/lxc 
 while its actually present under /cgroup.

I guess we should add the 'lxc' path to the cgroup in case of the 
ns_cgroup because the creation is handled by the kernel.

Serge ?




--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] nilfs

2012-03-02 Thread Serge Hallyn
Quoting Ulli Horlacher (frams...@rus.uni-stuttgart.de):
 On Fri 2012-03-02 (09:02), Daniel Baumann wrote:
 
  i'm not claiming btrfs is there yet, however, if you're using btrfs, you
  should at least make sure to use something remotely up2date, say 3.2.x.
 
 SLES11 SP2 was released this week with a 3.0 kernel and comes with btrfs.
 Same b(*CENSORED*)t as always from SuSE. What they label as Enterprise
 is Testing on Debian.

Some people have been testing btrfs on 3.1/3.2 kernels (in ubuntu
precise) with good results.

-serge

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] nilfs

2012-03-02 Thread John Drescher
 Some people have been testing btrfs on 3.1/3.2 kernels (in ubuntu
 precise) with good results.


I am using 3.1 / 3.2 kernels on 64 bit gentoo with btrfs at work on 2
production severs since ~ November of last year. One holds my lxc
containers for a samba bdc while the other container is a secondary
dns server.

John

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] adding a default gateway inside a container as a non root user

2012-03-02 Thread Serge Hallyn
Quoting Daniel Lezcano (daniel.lezc...@free.fr):
 On 03/02/2012 11:18 AM, Arun M wrote:
 
 
 With the 0.8.0 version, you will be able to setup the gateway directly
 from the configuration file.
 
 Cool. I wanted to try this so tested with the latest code from repository.
 
 However hit another issue now.
 
 $ lxc-execute -n alpha -f n1.conf -l DEBUG -o log -- /bin/busybox ash
 lxc-execute: No such file or directory - failed to rename cgroup 
 /cgroup//lxc/3784-/cgroup//lxc/alpha
 ...
 
 lxc-execute 1331137335.969 INFO lxc_cgroup - [1] found cgroup mounted at 
 '/cgroup',opts='rw,relatime,blkio,net_cls,freezer,devices,memory,cpuacct,cpu,ns,cpuset'
 lxc-execute 1331137335.969 DEBUGlxc_cgroup - get_init_cgroup: found init 
 cgroup for subsys (null) at /
 lxc-execute 1331137335.969 DEBUGlxc_cgroup - cgroup /cgroup has flags 0x1
 lxc-execute 1331137335.969 WARN lxc_cgroup - using deprecated ns_cgroup
 lxc-execute 1331137335.969 ERRORlxc_cgroup - No such file or directory - 
 failed to rename cgroup /cgroup//lxc/3840-/cgroup//lxc/alpha
 
 $ file /cgroup/3840
 /cgroup/3840: directory
 
 $ file /cgroup/lxc/3840
 /cgroup/lxc/3840: cannot open `/cgroup/lxc/3840' (No such file or directory)
 
 It appears that lxc-execute is trying to look for a dir under /cgroup/lxc 
 while its actually present under /cgroup.
 
 I guess we should add the 'lxc' path to the cgroup in case of the
 ns_cgroup because the creation is handled by the kernel.
 
 Serge ?

Drat, yes.

-serge

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users


Re: [Lxc-users] adding a default gateway inside a container as a non root user

2012-03-02 Thread Serge Hallyn
Quoting Bekir Dogan (beki...@gmail.com):
 Hi Daniel,
 
 On Thu, Mar 1, 2012 at 23:20, Daniel Lezcano daniel.lezc...@free.fr wrote:
  [...]
  With the 0.8.0 version, you will be able to setup the gateway directly
  from the configuration file.
 
 Can we see project plans or a todo list or something like these which
 gives us about the future features should come in next releases. I
 can't find something like this in a TODO file, open bug reports or a
 special page in sourceforge. Or do you suggest something to follow
 other than these.
 
 I'm about to start a project
 (https://github.com/bergerx/simplelxc#readme) for mainly laptop users
 to create and manage networking enabled test containers so simply
 which also setup initial networking for both lxc containers and host
 system.
 
 I would like to use networking as it should be and determine the
 project direction according to this or I should be in a position to
 re-implement some parts.
 
 I've talked implementing some ideas into debian lxc package Daniel
 Baumann mainy about a simple creation scenario without asking anything
 other than container name.

Right now in ubuntu precise, it's

lxc-create -t ubuntu -n p1
lxc-start -n p1

or

lxc-create -t ubuntu-cloud -n p1
lxc-start -n p1

If you want to deploy a bunch of cloned images, you can

lxc-create -t ubuntu -n plain -B lvm
 customize /dev/lxc/plain image if you like 
for i in `seq 1 20`; do
lxc-clone -s -o plain -n p$i
done

The only things different in Ubuntu are that a lxcbr0 bridge gets
installed by default (*1), and the release has been tweaked a bit so no
changes are needed to run in a container.  The first can't really
be done in lxc.git (because it's distro-dependent), and the second
is of course independent of lxc.

-serge

(*1) and if you don't specify a configuration file at creation time,
a default one using lxcbr0 is used.

--
Virtualization  Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing 
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
___
Lxc-users mailing list
Lxc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/lxc-users