Michael H. Warfield | 3 Mar 2009 18:52

Any prospect of a new 2.6.27 rpm?

Hey all (Hey Pavel),

	This is a two part question...

	I know everyone's busy with lots of things and I know cutting and
testing an rpm release is a lot of work, but...  I'm wondering if
there's a new 2.6.27 rpm waiting in the wings for release any time soon?
The last 2.6.27 (aivazovsky) had a very serious problem in the IPv6
firewall code that got quickly patched.  I had to abandon that one.  I
see there's been a lot of committs to that git branch that look pretty
serious as well.  Anything standing in the way?

	Now, outside of that, I do pull the git repository and I can build
things from that but I use rpm for managing the packages on my system
and would really rather avoid building straight from the git repository
and installing on the system.  On top of that, with rpm I can be sure
I've got all the other patches straight and all the build options
straight and it's much easier to copy between systems for testing on
other boxes.  I can't use the binary rpm's from the site anyways (due to
config option problems and due to the fact that they're built with a gcc
3.x compiler where all my Fedora systems are all on 4.x and I have other
external modules to compile where the compilers must agree) so I'm
always rebuilding the rpm's anyways.

	So...  What's the procedure (is there one) for generating the
appropriate "patch" to incorporate into the rpm build.  For the last
2.6.27 build, there was the patch-ovz-aivazovsky file which was then
rolled in with the original sources plus some configs and some other
patches into the rpmbuild/SOURCES directory for the rpmbuild.  How do I
go from the git repo and the base source tarball and derive that patch
(Continue reading)

Michael H. Warfield | 4 Mar 2009 02:31

Re: Any prospect of a new 2.6.27 rpm?

Hey All,

	I think I may have figured part of this (the second part) out for
myself.  In spite of being a kernel driver maintainer, I'm not a git
mavin by any means but I do slowly figure some things out.

On Tue, 2009-03-03 at 12:52 -0500, Michael H. Warfield wrote:
> Hey all (Hey Pavel),
> 
> 	This is a two part question...
> 
> 	I know everyone's busy with lots of things and I know cutting and
> testing an rpm release is a lot of work, but...  I'm wondering if
> there's a new 2.6.27 rpm waiting in the wings for release any time soon?
> The last 2.6.27 (aivazovsky) had a very serious problem in the IPv6
> firewall code that got quickly patched.  I had to abandon that one.  I
> see there's been a lot of committs to that git branch that look pretty
> serious as well.  Anything standing in the way?
> 
> 	Now, outside of that, I do pull the git repository and I can build
> things from that but I use rpm for managing the packages on my system
> and would really rather avoid building straight from the git repository
> and installing on the system.  On top of that, with rpm I can be sure
> I've got all the other patches straight and all the build options
> straight and it's much easier to copy between systems for testing on
> other boxes.  I can't use the binary rpm's from the site anyways (due to
> config option problems and due to the fact that they're built with a gcc
> 3.x compiler where all my Fedora systems are all on 4.x and I have other
> external modules to compile where the compilers must agree) so I'm
> always rebuilding the rpm's anyways.
(Continue reading)

Richard Ray | 5 Mar 2009 20:26
Picon
Favicon

vzctl devnodes help needed

If I use "vzctl set 101 --devnodes nst0:rw --save" it deletes any previous 
devnode setting in 101.conf.
Is it better to edit 101.conf without vzctl?
Should 101.conf have multiple DEVNODES lines, one for each device?
If HN has /dev/nst0 modes 666 and I have DEVNODES="nst0:rw" in 101.conf VE 
101 /dev/nst0 gets created with modes 644.
Can someone offer advice for handling devices in VE?

Thanks
Richard
Benoit Branciard | 5 Mar 2009 20:50
Picon
Favicon

vzctl 3.0.23: debian-add_ip.sh lenny (debian 5.0) IPv6 bug

the debian-add_ip.sh script available in vzctl 3.0.23 does not set up 
IPv6 default route correctly for a Debian 5.0 (Lenny) container.

The reason seems to be that the default behaviour of the ifconfig 
command changed from etch to lenny: when adding IPv6 addresses without 
specifying a prefixlen, the prefixlen is assumed to be /128 in Lenny 
instead of /0 in etch.

The patch below corrects the problem by explicitely specifying the 
prefixlen, and works for both Etch and Lenny containers.

--- debian-add_ip.sh.0  2009-03-05 18:57:40.000000000 +0100
+++ debian-add_ip.sh    2009-03-05 19:02:43.000000000 +0100
 <at>  <at>  -87,8 +87,8  <at>  <at> 
             printf "
  auto ${VENET_DEV}:${ifnum}
  iface ${VENET_DEV}:${ifnum} inet6 manual
-       up   ifconfig ${VENET_DEV}:${ifnum} add ${ip}
-       down ifconfig ${VENET_DEV}:${ifnum} del ${ip}
+       up   ifconfig ${VENET_DEV}:${ifnum} add ${ip}/0
+       down ifconfig ${VENET_DEV}:${ifnum} del ${ip}/0
  " >> ${CFGFILE}.bak
         fi

--

-- 
Ce message a ete verifie par MailScanner
pour des virus ou des polluriels et rien de
suspect n'a ete trouve.
Daniel Pittman | 6 Mar 2009 01:00
Gravatar

Re: vzctl devnodes help needed

Richard Ray <rray@...> writes:

> If I use "vzctl set 101 --devnodes nst0:rw --save" it deletes any
> previous devnode setting in 101.conf.  Is it better to edit 101.conf
> without vzctl?

Frequently, in my experience, but...

    vzctl set 101 --devnodes 'nst0:rw sg0:rw foobar:ro' --save

Regards,
        Daniel
Michael H. Warfield | 6 Mar 2009 16:36

Re: vzctl 3.0.23: debian-add_ip.sh lenny (debian 5.0) IPv6 bug

On Thu, 2009-03-05 at 20:50 +0100, Benoit Branciard wrote:
> the debian-add_ip.sh script available in vzctl 3.0.23 does not set up 
> IPv6 default route correctly for a Debian 5.0 (Lenny) container.

> The reason seems to be that the default behaviour of the ifconfig 
> command changed from etch to lenny: when adding IPv6 addresses without 
> specifying a prefixlen, the prefixlen is assumed to be /128 in Lenny 
> instead of /0 in etch.

	That doesn't make sense to me.  I would think the /0 (which should be
all addresses are valid to the local subnet) would be, should be, wrong.
Looking at my systems, I don't see a single /0 address anywhere.
Loopback has a specific ::1/128 address but then all other addresses
are /64's including the link local addresses.  Most of my global unicast
addresses and routes are assigned via stateless autoconfig but where I
have static assignments for servers, they're all /64 (which is what a
standard subnet is suppose to be).  Some of my tunnels are either /128
or /112 depending on the specifics of the PtoP tunnel convention (varies
with broker).  That's normal standard IPv6 behavior that's expected.

	I see nothing where a /0 would be valid.  I find it weird that Etch
would have been setting it up that way.  That doesn't sound valid in the
first place.  Maybe something peculiar to the vnet interfaces, which
I've been avoiding because of non-standard IPv6 behavior.

	All my stuff is Fedora, CentOS, or Ubuntu on hard iron and only Fedora
and CentOS in OpenVZ containers.  All the containers are using veth
devices and stateless autoconfig.

> The patch below corrects the problem by explicitely specifying the 
(Continue reading)

Benoit Branciard | 6 Mar 2009 18:53
Picon
Favicon

Re: vzctl 3.0.23: debian-add_ip.sh lenny (debian 5.0) IPv6 bug

Michael H. Warfield a écrit :
> On Thu, 2009-03-05 at 20:50 +0100, Benoit Branciard wrote:
>> the debian-add_ip.sh script available in vzctl 3.0.23 does not set up 
>> IPv6 default route correctly for a Debian 5.0 (Lenny) container.
> 
>> The reason seems to be that the default behaviour of the ifconfig 
>> command changed from etch to lenny: when adding IPv6 addresses without 
>> specifying a prefixlen, the prefixlen is assumed to be /128 in Lenny 
>> instead of /0 in etch.
> 
> 	That doesn't make sense to me.  I would think the /0 (which should be
> all addresses are valid to the local subnet) would be, should be, wrong.

On a normal ethernet device yes, /0 is wrong.
On a venet device /0 is needed to force default route to be forwarded to 
the host. This seems messy, but this is the way it works.

> 	I see nothing where a /0 would be valid.  I find it weird that Etch
> would have been setting it up that way.

Yes, having ifconfig defaulting to /0 seems like a bug in Etch, which 
has been corrected in Lenny. /128 is a far more acceptable default value 
(that is, a single IP address).

But having debian-add_ip.sh NOT specifying the REQUIRED /0 (in venet 
context), and relying on a buggy default value, is another bug.

> 
>> The patch below corrects the problem by explicitely specifying the 
>> prefixlen, and works for both Etch and Lenny containers.
(Continue reading)

unni krishnan | 8 Mar 2009 06:51
Picon
Gravatar

Kernel warning/error when booting the kernel

Hi,

I am getting a number of following messages when booting the OpenVZ kernel

[root <at> localhost ~]# grep NOHZ /var/log/messages
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 20
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 20
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 20
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
(Continue reading)

Sterling Windmill | 8 Mar 2009 20:10

Re: Kernel warning/error when booting the kernel

It may be a kernel bug that was fixed by this patch:

http://lkml.org/lkml/2007/5/22/35

Regardless, I wouldn't recommend using the 2.6.22 OpenVZ kernel as it hasn't been touched since late 2007 and is no longer maintained.

----- Original Message -----
From: "unni krishnan" <unnikrishnan.a-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
To: users-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org
Sent: Sunday, March 8, 2009 12:51:03 AM GMT -05:00 US/Canada Eastern
Subject: [Users] Kernel warning/error when booting the kernel

Hi,

I am getting a number of following messages when booting the OpenVZ kernel

[root <at> localhost ~]# grep NOHZ /var/log/messages
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 20
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:30:20 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 20
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 00:32:45 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 20
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10
Mar  6 11:23:55 localhost kernel: NOHZ: local_softirq_pending 10



But I am able to get the system up with that kernel.

[root <at> localhost ~]# uname -a
Linux localhost.localdomain 2.6.22-ovz005 #1 SMP Fri Mar 6 00:21:04
IST 2009 i686 athlon i386 GNU/Linux
[root <at> localhost ~]#



Do you have any idea ?

--
---------------------
With regards,
Unni

"A candle loses nothing by lighting another candle"
_______________________________________________
Users mailing list
Users-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org
https://openvz.org/mailman/listinfo/users
_______________________________________________
Users mailing list
Users@...
https://openvz.org/mailman/listinfo/users
Andrew Boag | 10 Mar 2009 13:55
Picon

Directory structure of VE on host

I'm running openvz on an ubuntu 8.0.6 server box. Working well and I'm 
really enjoying being able to script up new VE's when I need them.

I'm just trying to figure out the contents of /var/lib/vz (where I'm 
storing the VE's)

I see 'private' and 'root' which both look like the file system of the 
VE. (they contain the VE ID number, then what looks like the file system 
of the VE).

If I wanted to back up the file system of the VE (happy to stop the VE 
while the happens), what should I back up???

Or can someone explain the difference between private and root?

Cheers

Gmane