OpenVZ Forum


Home » General » Support » Problem with Networking - VENET Interfaces not Configured
Problem with Networking - VENET Interfaces not Configured [message #30715] Wed, 04 June 2008 16:47 Go to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
Hi,

Im having a few problems with networking.

This is what I see. The VPS venet0 interface doesn't have an IP, and the venet0 doesn't appear to be bridged to the ethernet card on eth1.

The VE doesn't seem to bring up venet0 or venet0:0 (indeed why I have two - Im not sure).

Trying to restart the network services on the VE from the VPS gives:

"Bringing up interface venet0: Error, some other host already uses address 127.0.0.1."

venet0 seems assigned 127.0.0.1 while venet0:0 10.0.0.32 (my chosen ip), which means if it can't bring up venet0 then I would imagine it can't load venet0:0

Even if I change the IP of venet0 on the VE to something unused, it still gives the same error.

What this means, is that I can't ping the IP of 10.0.0.32 from either the network or the VPS.

Here's more info:

Kernel
Quote:

Linux rutsey.blargle.local 2.6.24-ovz005.1 #1 SMP Mon May 12 16:38:09 MSD 2008 i 686 athlon i386 GNU/Linux


Setting IP Address With VZCTL
Quote:

Adding IP address(es): 10.0.0.32
SIOCADDRT: No such device
Saved parameters for VE 102


VPS ifcofnig
Quote:

venet0 Link encap:UNSPEC HWaddr 00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00
UP BROADCAST POINTOPOINT RUNNING NOARP MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)

VPS Route
Destination Gateway Genmask Flags Metric Ref Use Iface
10.0.0.32 * 255.255.255.255 UH 0 0 0 venet0


VE Scripts and Files
Quote:

[root@rutsey sysconfig]# cat network
NETWORKING="yes"
GATEWAY="192.0.2.1"
HOSTNAME="collins.blargle.local"


lo on ve
Quote:

[root@rutsey network-scripts]# cat ifcfg-lo
DEVICE=lo
IPADDR=127.0.0.1
NETMASK=255.0.0.0
NETWORK=127.0.0.0
# If you're having problems with gated making 127.0.0.0/8 a martian,
# you can change this to something else (255.255.255.255, for example)
BROADCAST=127.255.255.255
ONBOOT=yes
NAME=loopback


venet0 on ve
Quote:

[root@rutsey network-scripts]# cat ifcfg-venet0
DEVICE=venet0
BOOTPROTO=static
ONBOOT=yes
IPADDR=127.0.0.1
NETMASK=255.255.255.255
BROADCAST=0.0.0.0


venet0:0 on ve
Quote:

[root@rutsey network-scripts]# cat ifcfg-venet0:0
DEVICE=venet0:0
ONBOOT=yes
IPADDR=10.0.0.32
NETMASK=255.255.255.255


Network Stop
Quote:

[root@rutsey etc]# vzctl exec 102 service network stop
Shutting down loopback interface: [ OK ]
Cannot find device "sit0"


Network Start
Quote:

[root@rutsey etc]# vzctl exec 102 service network start
Bringing up loopback interface: Error, some other host already uses address 127.0.0.1.
[FAILED]
Bringing up interface venet0: Error, some other host already uses address 127.0.0.1.
[FAILED]
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists
RTNETLINK answers: File exists


Let me know if you need more.



Im trying the fedora-9-i386 template.

Any thoughts?

Cheers
Simon
Re: Problem with Networking - VENET Interfaces not Configured [message #30716 is a reply to message #30715] Wed, 04 June 2008 18:51 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
Looks like it was the guest fedora-9. Went back to fedora-6.

That's kind of putting me off to be honest.
Re: Problem with Networking - VENET Interfaces not Configured [message #30721 is a reply to message #30716] Wed, 04 June 2008 22:04 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
I must be doing something wrong. Maybe I havent found the right documentation, but I've seen people who've created their own containers for Smoothwall, people must have got Fedora 7, 8 and 9 working otherwise there wouldn't be any templates.

Someone, please, tell me what Im doing wrong? If I could build a Smoothwall 3.0 container, I'd switch! But I've followed the instructions as much as I can before I start configuring ethernet devices but the container keeps shutting down on me!
Re: Problem with Networking - VENET Interfaces not Configured [message #30732 is a reply to message #30715] Thu, 05 June 2008 08:03 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hi,

don't worry I think I can reproduce that issue. Now I'll try to investigate it.
Re: Problem with Networking - VENET Interfaces not Configured [message #30736 is a reply to message #30721] Thu, 05 June 2008 09:25 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hi,

is it very important for you to use OpenVZ 2.6.24 based kernel?
Please, try to do the same with 2.6.18 or RHEL based kernels.
Re: Problem with Networking - VENET Interfaces not Configured [message #30743 is a reply to message #30736] Thu, 05 June 2008 11:51 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
There isn't any particular reason why I need to use the 2.6.24 kernel. I just tried 2.6.18 first on my Fedora 9 x64 distro and it failed miserably.

It panicked very quickly with:

"not syncing: attempted to kill init"

in relation to something along the lines of.... "netlink_update_listeners"

I tried rebuilding the kernel, including from vanilla .18 but had a problem with sumversion not having an include to limits.h.

Trying to install .18 kernel rpm on my current Fedora 8 i386, I get a warning about a non-existing module, I think pata_amd.

I think if I must build an .18 kernel, I'll go from vanilla source again and apply the patches. I don't get on at all well with the source rpm from OpenVZ.

I'll give that a try and let you know.

Cheers
Simon
Re: Problem with Networking - VENET Interfaces not Configured [message #30744 is a reply to message #30743] Thu, 05 June 2008 12:21 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hi,

Quote:


"not syncing: attempted to kill init"


May be it's related to the workaround described here http://wiki.openvz.org/Modifying_initrd_image

Quote:


Trying to install .18 kernel rpm on my current Fedora 8 i386, I get a warning about a non-existing module, I think pata_amd.



Are you sure that that module is not compiled to kernel?
Re: Problem with Networking - VENET Interfaces not Configured [message #30814 is a reply to message #30744] Sat, 07 June 2008 12:17 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
Thanks for the reply.

2.6.18 definitely works in Fed 8 i386. I personally think Fed 9 x64 is a bit ahead of the game (read Buggy).

Im fairly happy to continue with Fed 8 i386 and OpenVZ.

Truth be told I've just tried Xen but I can't for the life of me, use pciback to get a device forwarded to the guest OS. It'll be captured by the pciback driver, but won't forward to the guest OS. I thought maybe it was because a standard Fedora kernel didn't have the extensions for the front end drivers to capture the PCI card. So I went an installed kernel-xen into the guest OS and now it won't boot!

Just tried re-installing again, and same results!

Two bad things I thought of with Xen and the like.

Firstly migration means if you're using a file, transferring a 20 Gig file to another host!

Secondly, the problems with hardware support.

Thirdly, if something goes wrong with the file, you're stuffed!

So Im back in the OpenVZ camp.

Im happy to run the OpenVZ provided guest OS's.

But I need to find out how to debug my own template. That's the main thing for me. If I try to start the container it starts ish, but shuts down quickly and I can't find the debugging information.

Any thoughts?

Cya
Simon
Re: Problem with Networking - VENET Interfaces not Configured [message #30815 is a reply to message #30814] Sat, 07 June 2008 12:36 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hello,

Quote:


I can't find the debugging information



please describe in more detail what does it mean?
What kind of debug information would you like to see?
Re: Problem with Networking - VENET Interfaces not Configured [message #30817 is a reply to message #30815] Sat, 07 June 2008 12:44 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
Something about why the guest failed to boot... Something that would give me an idea of where to look at.

For example, quite literally, within two seconds of trying to start my Virtual guest:

[root@rutsey /]# vzctl status 101
VEID 101 exist mounted running
[root@rutsey /]# vzctl status 101
VEID 101 exist mounted down
[root@rutsey /]#
Re: Problem with Networking - VENET Interfaces not Configured [message #30818 is a reply to message #30817] Sat, 07 June 2008 12:55 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hi,

anything in /var/log/vzctl.log, /var/log/messages, dmesg?
What kind of kernel do you use (the exact version!)? Where have you got it? What kind of template do you use? Where have you got it?
You have mentioned about debugging information. What kind of information have you prepared?
Re: Problem with Networking - VENET Interfaces not Configured [message #30821 is a reply to message #30818] Sat, 07 June 2008 13:27 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
/var/lot/messages:

Jun  7 13:40:46 rutsey kernel: CT: 101: started
Jun  7 13:40:46 rutsey kernel: CT: 102: started
Jun  7 13:40:48 rutsey kernel: CT: 103: started
Jun  7 13:40:49 rutsey kernel: CT: 101: stopped
Jun  7 13:40:56 rutsey kernel: CT: 101: started
Jun  7 13:40:59 rutsey kernel: CT: 101: stopped



(Problematic Container is 101)

Here is some further information from /var/log/vzctl.log:

2008-06-07T14:22:49+0100 vzctl : Setting CPU units: 1000
2008-06-07T14:22:49+0100 vzctl : WARNING: Settings were not saved and will be resetted to original values on next start (use --save flag)
2008-06-07T14:22:49+0100 vzctl : VE 101 : Starting VE ...
2008-06-07T14:22:49+0100 vzctl : VE 101 : VE is mounted
2008-06-07T14:22:49+0100 vzctl : VE 101 : Adding IP address(es): 10.0.0.34
2008-06-07T14:22:50+0100 vzctl : VE 101 : Warning: add_ip action script is not specified
2008-06-07T14:22:50+0100 vzctl : VE 101 : Setting CPU units: 1000
2008-06-07T14:22:50+0100 vzctl : VE 101 : Configure meminfo: 65536
2008-06-07T14:22:50+0100 vzctl : VE 101 : Warning: set_hostname action script is not specified
2008-06-07T14:22:50+0100 vzctl : VE 101 : Warning: set_dns action script is not specified
2008-06-07T14:22:50+0100 vzctl : VE 101 : VE start in progress...
2008-06-07T14:22:50+0100 vzctl : VE 102 : Starting VE ...
2008-06-07T14:22:50+0100 vzctl : VE 102 : VE is mounted
2008-06-07T14:22:50+0100 vzctl : VE 102 : Adding IP address(es): 10.0.0.32
2008-06-07T14:22:51+0100 vzctl : VE 102 : Setting CPU units: 1000
2008-06-07T14:22:51+0100 vzctl : VE 102 : Configure meminfo: 65536
2008-06-07T14:22:51+0100 vzctl : VE 102 : Set hostname: lifeson.blargle.local
2008-06-07T14:22:51+0100 vzctl : VE 102 : File resolv.conf was modified
2008-06-07T14:22:51+0100 vzctl : VE 102 : VE start in progress...
2008-06-07T14:22:51+0100 vzctl : VE 103 : Starting VE ...
2008-06-07T14:22:51+0100 vzctl : VE 103 : VE is mounted
2008-06-07T14:22:51+0100 vzctl : VE 103 : Adding IP address(es): 10.0.0.31
2008-06-07T14:22:52+0100 vzctl : VE 103 : Setting CPU units: 1000
2008-06-07T14:22:52+0100 vzctl : VE 103 : Configure meminfo: 65536
2008-06-07T14:22:52+0100 vzctl : VE 103 : Set hostname: peart.blargle.local
2008-06-07T14:22:52+0100 vzctl : VE 103 : File resolv.conf was modified
2008-06-07T14:22:52+0100 vzctl : VE 103 : VE start in progress...
2008-06-07T14:23:05+0100 vzctl : VE 101 : Starting VE ...
2008-06-07T14:23:05+0100 vzctl : VE 101 : VE is unmounted
2008-06-07T14:23:05+0100 vzctl : VE 101 : VE is mounted
2008-06-07T14:23:05+0100 vzctl : VE 101 : Adding IP address(es): 10.0.0.34
2008-06-07T14:23:05+0100 vzctl : VE 101 : Warning: add_ip action script is not specified
2008-06-07T14:23:05+0100 vzctl : VE 101 : Setting CPU units: 1000
2008-06-07T14:23:05+0100 vzctl : VE 101 : Configure meminfo: 65536
2008-06-07T14:23:05+0100 vzctl : VE 101 : Warning: set_hostname action script is not specified
2008-06-07T14:23:05+0100 vzctl : VE 101 : Warning: set_dns action script is not specified
2008-06-07T14:23:05+0100 vzctl : VE 101 : VE start in progress...


I'd be happy to get anything else you need, but I havent a clue where Im searching (that's why I ask if it's possible to increase the debugging level...).

But let me know what you need, and I'll provide.

To recap again though:

Fedora Core 8 i686 host on an AMD Athlon 64 3500 system, guest os is supposed to be Smoothwall 3.0. I took a tar of this from a system installed on KVM/Qemu and used this as my template.

I've edited as much as I can, following the instructions regarding to create a VE from a Physical Host. Perhaps there's something I've left out, but Im guessing for it to go down as quickly as it does and not log any messages in its own area, there must be something very significantly wrong. Remember, this particular guest goes down in about 2 seconds.

Cheers
Simon

[Updated on: Sat, 07 June 2008 13:29]

Report message to a moderator

Re: Problem with Networking - VENET Interfaces not Configured [message #30824 is a reply to message #30821] Sat, 07 June 2008 14:10 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Quote:


2008-06-07T14:23:05+0100 vzctl : VE 101 : Starting VE ...
2008-06-07T14:23:05+0100 vzctl : VE 101 : VE is unmounted
2008-06-07T14:23:05+0100 vzctl : VE 101 : VE is mounted
2008-06-07T14:23:05+0100 vzctl : VE 101 : Adding IP address(es): 10.0.0.34
2008-06-07T14:23:05+0100 vzctl : VE 101 : Warning: add_ip action script is not specified
2008-06-07T14:23:05+0100 vzctl : VE 101 : Setting CPU units: 1000
2008-06-07T14:23:05+0100 vzctl : VE 101 : Configure meminfo: 65536
2008-06-07T14:23:05+0100 vzctl : VE 101 : Warning: set_hostname action script is not specified
2008-06-07T14:23:05+0100 vzctl : VE 101 : Warning: set_dns action script is not specified
2008-06-07T14:23:05+0100 vzctl : VE 101 : VE start in progress...




Sorry, but I cannot find where is VE was stopped?
Anyway, you haven't provide with the exact kernel version.
There are some parameters in /etc/vz/vz.conf file (man vz.conf) that increase log level. (LOG_LEVEL, VERBOSE).
Could you possibly give me VEs tarball and conf file of that VE I'll try to do the same.
Re: Problem with Networking - VENET Interfaces not Configured [message #30825 is a reply to message #30824] Sat, 07 June 2008 14:20 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
that's it though, it doesn't log that it's stopped - but vzctl status 101 reports that that it's mounted down and it doesn't show up in vzlist.

I've increased logging and verbosity, but again, it just says the VE is starting then it mounts down with no information.

I'd have to get back to you with the kernel information, but can provide both original and current tarball.... Let me know where to send it.

Thanks
Simon
Re: Problem with Networking - VENET Interfaces not Configured [message #30828 is a reply to message #30825] Sat, 07 June 2008 14:32 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Could you possibly provide the link where I can get that template (via PM)?
Unfortunately, for me there is only this way to get it?
Re: Problem with Networking - VENET Interfaces not Configured [message #30829 is a reply to message #30828] Sat, 07 June 2008 14:51 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
Ok, I'll do that.

Just to let you know, the Smoothwall Kernel is 2.6.16.53. Im going to actually just take a snapshot of a fresh install I've done and see if I get better luck with that.

I'll get you two copies - the one I stripped down trying to follow the instruction on creating a VE from a Physical device, and the other which will just be the entire image untouched - apart from removing proc and dev etc.

The kernels I have been using with same results are:
ovzkernel-2.6.18-53.1.19.el5.028stab053.14.i686.rpm (RHEL5)
kernel-2.6.24-ovz005.1.i686.rpm

Ta
Simon
Re: Problem with Networking - VENET Interfaces not Configured [message #30837 is a reply to message #30828] Sun, 08 June 2008 13:09 Go to previous messageGo to next message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
I've got a little more info by the way.

If I try to do a vzctl enter immediately after a vzctl start, I get:

Unable to get tty name: Bad file descriptor
smoothwall (root) / $ got signal 9
exited from VE 104

Performing an strace, I find:

[pid 22793] readlink("/proc/self/fd/5", 0x92f9fa0, 4095) = -1 ENOENT (No such file or directory)
[pid 22793] write(2, "Unable to get tty name", 22) = 22
[pid 22793] write(2, ": Bad file descriptor", 21) = 21
[pid 22793] write(2, "\n", 1)           = 1
[pid 22793] ioctl(5, SNDCTL_TMR_TIMEBASE or TCGETS, {B38400 opost isig icanon echo ...}) = 0
[pid 22793] readlink("/proc/self/fd/5", 0x92f9fa0, 4095) = -1 ENOENT (No such file or directory)


There were a few devices missing from 104, console, core, fd, full, kmem, mem and port. I've generated these as best I could.

And here's some output after I managed to perform a very quick ls -al on /proc inside the VE:

smoothwall (root) / $ ls -al /proc
total 8
dr-xr-xr-x  2 root root 4096 Jun  8 13:32 ./
drwxr-xr-x 16 root root 4096 Jun  7 17:36 ../


And finally, if I try ps ax:

Error, do this: mount -t proc none /proc


Maybe there's something up with the proc folder?

Nearly got those files ready.

they work out at about 60 Meg each.

Cheers
Simon
Re: Problem with Networking - VENET Interfaces not Configured [message #30840 is a reply to message #30828] Sun, 08 June 2008 21:36 Go to previous message
sjdean is currently offline  sjdean
Messages: 30
Registered: May 2008
Member
I did some chobbling around with rc.sysinit on the guest for a bit of fun, and copied some bits from a Working Fedora 8 template I put together.

And it works now.

I think it was down to /proc not being mounted correctly.

That's resolved. Just a bit of a shame at the moment that the software doesn't seem to like me using venet0:0 devices - doesn't even show up in its screens.

Im not saying that's a flaw, and indeed, Im sure I could probably rewrite bits of their code if I can find the time.

Indeed, there's a file called interfaces.cgi which readS:

sub display_interface
{
        my ( $settings, $prefix ) = @_;

        my $interface = $settings{"${prefix}_DEV"};

       return if ($interface !~ /eth[0123]/ );

        my $ifconfig_details = &pipeopen('/sbin/ifconfig', $interface);
.
.
.
.



Place a # before that return and it brings out the details for the interface. What a load of waffle. Bad bad bad bad bad hard coded rubbish. I wonder where else there might be hard coded stuff?

Cheers
Simon
Previous Topic: How do I reset root password.
Next Topic: SSH proxy/gateway
Goto Forum:
  


Current Time: Thu Apr 25 08:20:21 GMT 2024

Total time taken to generate the page: 0.01587 seconds