Brain Fallik, this bug was reported a while ago and there hasn't been
any activity in it recently. We were wondering if this is still an
issue? If so, could you please test for this with the latest server
release of Ubuntu? ISO images are available from
http://releases.ubuntu.com/raring/ .
If it
** Changed in: bridge-utils (Ubuntu)
Status: Confirmed = Invalid
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
https://bugs.launchpad.net/bugs/616064
Title:
Broken networking in kvm guests with pcnet
** Summary changed:
- Broken networking in kvm guests
+ Broken networking in kvm guests with pcnet driver
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
https://bugs.launchpad.net/bugs/616064
Title:
Broken
I have the same problem on 11.10 (old version had it too), if I scp/nfs
a large file it starts, go fast but at some point the network of the
machine does not work anymore. sometimes it transfers 1GB others a
100MB. I use the bridge interface to get to the machine and it still
responding, it seems
Following the advice on
http://serverfault.com/questions/341178/downloads-stuck-from-a-vm-run-by-libvirt-kvm-on-ubuntu-10-04-lts
I changed:
model type='pcnet'/
To
model type='rtl8139'/
And now I'm able to transfer my 37GB tar, so it seems a bug in the
virtual drivers.
BTW It's not only
@Daniel: yes, you can use tap devices with kvm, and attach the tap
devices to the bridge.
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
https://bugs.launchpad.net/bugs/616064
Title:
Broken networking in kvm
Interesting. If multiple downloads of small files is the problem, could
you retry the experiment from comment #46 but using the small wgets?
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
1.
$ virsh dumpxml vs-maven
domain type='kvm' id='4'
namevs-maven/name
uuid40f97e28-afc2-1730-3dfb-292074bae1d6/uuid
memory2097152/memory
currentMemory2097152/currentMemory
vcpu1/vcpu
os
type arch='x86_64' machine='pc-0.12'hvm/type
boot dev='hd'/
/os
features
acpi/
Do you want me to test another mechanism ?
Thanks, but no thanks. While it's not 100% conclusive, this certainly
suggests that the bug may be in the driver in qemu.
To help try to reproduce, can you give:
1. output of 'virsh dumpxml vm-name'
2. the release of both your host and guest.
Here is what I done in my client :
user@host:~$ cat /dev/urandom | hexdump | head -c 1048576 /tmp/1MB.txt
user@host:~$ N=1; while [ 1 ]; do echo $N | nc 192.168.1.214 8080; cat
/tmp/1MB.txt | nc 192.168.1.214 8080; ((N += 1)); sleep 0.5s; done
No problem here, in terminal 1 all was displayed
To reproduce the use case, I need that nc serves big files from the
server and that it would be my client which would download the 1MB
file... but I don't manage to use nc correctly to do that.
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is
If all goes well the instructions in comment #35 will cause no problems
and the VM will remain reachable. However a typo could indeed endanger
the network and be less than trivial to fix.
The experiment would only help us to determine whether the bug comes
down to the kernel or part of the
Is there any ywireshark data showing arp and dns traffic at the same
time as the conversation you posted? (it's not in the file itself)
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
Well I've tried this (after first steps with git)
# terminal 1
root@vs-master:~# ip link add type veth
# terminal 2
root@vs-master:~# /bin/ns_exec -cmn /bin/bash
about to clone with 4002
clone: Invalid argument
root@vs-master:~# echo $$
7934
# terminal 1
root@vs-master:~# ifconfig veth0
It's just another bridge, so you should be able to substitude br0 for
virbr1. Note that if veth0 and veth1 already exist from your last test,
you can either re-use them, or else when you do another 'ip link add
type veth' you'll create veth2 and veth3 and can use those.
--
You received this bug
I've tested
# terminal 1
root@vs-master:~# brctl delif virbr1 veth0
root@vs-master:~# brctl addif br0 veth0
root@vs-master:~# ip link set veth1 netns 5039
root@vs-master:~# nc -4 -k -l 8080
# terminal 2
root@vs-master:~# /bin/ns_exec -cmn /bin/bash
about to clone with 4002
clone: Invalid
Could you try wget of large files instead of the echo?
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
https://bugs.launchpad.net/bugs/616064
Title:
Broken networking in kvm guests
To manage notifications
Unfortunately i don't see anything telling in the wireshark log.
Could you try out the instructions in comment #33 (taking into account
the corrections in comment #35)?
When a VM is running, what do
ifconfig -a
netstat -nr
show?
** Changed in: qemu-kvm (Ubuntu)
Assignee: Serge
Thanks for the info, Anthony. The wireshark info failed to download for
me. Could you attach it here?
My only guess ATM is that it is a bad interaction with the physical
switch or router to which eth3 is attached. I'd ask you to try with
different hardware, but I see you're running production
** Changed in: linux (Ubuntu)
Status: New = Confirmed
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
https://bugs.launchpad.net/bugs/616064
Title:
Broken networking in kvm guests
To manage
Yes, here is the Wireshark capture.
Actually, the eth3 is leading to a NetAsq U70-A firewall (8.0.3
firmware) on a DMZ subnetwork. The client I'm using is located on
another subnetwork attached to that firewall.
I can make some test if you want and send me directives...
Moreover, the server has
marking 'confirmed' in bridge-utils just to get this off new bug
lists.
** Changed in: bridge-utils (Ubuntu)
Status: New = Confirmed
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to bridge-utils in Ubuntu.
Note I made two mistakes in comment 33:
1. only do the 'ip link add type veth' once, not twice,
and
2. the nc sender should send to port , not 999, to match the receiver
port, of course.
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug
Serge,
Apologies for the delay. This bug occurs on a production system for us
so it's challenging to find time to run experiments. Now is a good time
so I should have some new results for you shortly.
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this
Serge,
I've completed (I hope) your suggestions. Unfortunately neither the new
KVM nor the new kernel seemed to resolve the issue. More details below
but I'm not sure where to take this from here. As you point out, it
might be an issue with our environment. I doubt it's bad software
installed
Quoting Brian Fallik (616...@bugs.launchpad.net):
6. apt-add-repository ppa:kernel-ppa/pre-proposed
Did you do an apt-get update here? If not, then you didn't get the
updated kernels.
7. apt-get install linux-headers-server linux-image-server linux-server
8. rebooted the machine
9.
Serge,
On Fri, Sep 10, 2010 at 12:56 PM, Serge Hallyn
616...@bugs.launchpad.net wrote:
Quoting Brian Fallik (616...@bugs.launchpad.net):
6. apt-add-repository ppa:kernel-ppa/pre-proposed
Did you do an apt-get update here? If not, then you didn't get the
updated kernels.
Yes, I just missed
Addendum to the above - it's been pointed out to me that scp DOES trigger the
bug for you, so I was completely unable to reproduce your bug.
My setup however was using the 0.12.5 kvm listed above, as well as the
lucid pre-proposed kernels from
Alas, the unfirewalled machine I was going to try on has now been
taken away. I may be able to reproduce this using user-space qemu
on a VM on a cluster, and will try that next week if I have to, but
in the meantime:
I'm planning to try to upload qemu-kvm 0.12.5 to the lucid-backports
ppa in the
(Note to self - it may be worth-retrying with an i386 guest on previous
setup)
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to qemu-kvm in ubuntu.
--
Serge,
I switched networking back to NAT and repeated the svn co to verify the
problem still exists. It does. Then I tried the wget operation 5 times
and all of them succeeded.
I also verified that we're using the same default.xml configuration. They're
almost identical except for one
I switched networking back to NAT and repeated the svn co to verify the
problem still exists. It does. Then I tried the wget operation 5 times
and all of them succeeded.
Ok, then I'll have to re-test on a machine where I'm not firewalled.
Will post the results as soon as I have a chance to set
On Mon, Aug 30, 2010 at 2:04 PM, Serge Hallyn 616...@bugs.launchpad.net wrote:
Side question: every time I switch networking times, the guest bumps
ethernet device number (eth0, eth1, ...). This makes it annoying to
reset networking to switch static/dhcp on the next reboot. Is there a
method
Brian, do you get the bug also when you just wget a large file (say
wget
http://archive.ubuntu.com/ubuntu/pool/restricted/b/bcmwl/bcmwl_5.60.48.36+bdcom.orig.tar.gz
)?
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a
I set up a NAT config using the following
/etc/libvirt/qemu/networks/default.xml:
network
namedefault/name
bridge name=virbr%d /
forward mode=nat/
ip address=192.168.122.1 netmask=255.255.255.0
dhcp
range start=192.168.122.2 end=192.168.122.254 /
/dhcp
/ip
/network
but
Hi Brian,
I think the two easiest things to test next are:
1. in my setup, stp was disabled on br0, in yours I think it was
enabled. You might try your setup with stp disabled (which really
should not be what you want, but it's a test).
2. you said
In the scenario you suggested which seemed
On Thu, Aug 26, 2010 at 9:01 AM, Serge Hallyn 616...@bugs.launchpad.net wrote:
2. you said
In the scenario you suggested which seemed to work, tap1 was bridged
directly to eth0 and used an IP address from the host's actual subnet. This
is similar to:
Serge,
I've gotten the bridge configuration set up and everything seems stable
and working after a few hours of testing. Obviously we'll keep watching
it over the next few days. Thanks for the suggestion for the
workaround. It'd be nice to understand the issue with the original
configuration
Quoting Brian Fallik (616...@bugs.launchpad.net):
Serge,
I've gotten the bridge configuration set up and everything seems stable
and working after a few hours of testing. Obviously we'll keep watching
it over the next few days. Thanks for the suggestion for the
workaround. It'd be nice
Serge,
$ brctl show
bridge name bridge id STP enabled interfaces
br0 8000.001a64b33b10 no eth0
vnet0
vnet1
virbr0
can you try 'brctl stp br0 on' and see if all still works?
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to qemu-kvm in ubuntu.
--
Ubuntu-server-bugs mailing list
Thanks, Brian - no, next it's up to me to try to reproduce with the
NAT config. I need to rebuilt my test system first though. Hoping
to have results in the afternoon.
** Changed in: qemu-kvm (Ubuntu)
Assignee: (unassigned) = Serge Hallyn (serge-hallyn)
--
Broken networking in kvm guests
I failed to reproduce this. Here is how I tried. I have a
10.04.1 lucid host. My /etc/network/interfaces looks like:
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).
# The loopback network interface
auto
Quoting Brian Fallik (616...@bugs.launchpad.net):
Serge - my uploads in comment #5 and comment #6 should have been while
guest #1 (Lucid) was running. This problem occurs regardless of 1 or 2
guests are running. I've stopped launching the second guest (Hardy)
since the number of guests
Tried to reproduce with a hardy guest, but that failed as well.
Did this use to work before a particular upgrade, or is this the first
you've tried?
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Serge,
I think I'm starting to understand this issue and I'm fairly certain the
problem is in libvirt.
First off, it seems like I can repeatedly checkout depot_tools in a KVM
using a setup that I think matches your example except for a new (and
different) error.
My first step was shutdown
Brian,
Can you please include the info from:
brctl show
ifconfig -a
Both those commands should be executed on the host.
Thanks!
** Changed in: qemu-kvm (Ubuntu)
Status: Confirmed = Incomplete
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You
** Attachment added: brctl show
https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/616064/+attachment/1516198/+files/brctl_show.log
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Server
** Attachment added: ifconfig -a
https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/616064/+attachment/1516203/+files/ifconfig_a.log
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Server
** Changed in: qemu-kvm (Ubuntu)
Status: Incomplete = Confirmed
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to qemu-kvm in ubuntu.
--
Ubuntu-server-bugs
Re: I can see these RST packets via Wireshark but can't explain their
source. Where do those RST packets appear to come from ?
** Changed in: qemu-kvm (Ubuntu)
Importance: Undecided = Medium
** Changed in: qemu-kvm (Ubuntu)
Status: New = Incomplete
--
Broken networking in kvm guests
Here's example failure output from running the command inside the Lucid guest:
$ svn co http://src.chromium.org/svn/trunk/tools/depot_tools
svn: PROPFIND of '/svn/trunk/tools/depot_tools': Could not read status line:
Connection reset by peer (http://src.chromium.org)
The same command is
** Changed in: qemu-kvm (Ubuntu)
Status: Incomplete = Confirmed
--
Broken networking in kvm guests
https://bugs.launchpad.net/bugs/616064
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to qemu-kvm in ubuntu.
--
Ubuntu-server-bugs
** Attachment added: BootDmesg.txt
http://launchpadlibrarian.net/53444870/BootDmesg.txt
** Attachment added: CurrentDmesg.txt
http://launchpadlibrarian.net/53444871/CurrentDmesg.txt
** Attachment added: Dependencies.txt
http://launchpadlibrarian.net/53444872/Dependencies.txt
**
54 matches
Mail list logo
