Dear Peter,
I also rename the veth to the name of the container and without this, as the
default a random name will be used. You'll see why with the later you don't
notice an issue:
If a container goes down, the veth did not vanish as long as there is a tcp
connection using it. And in case of
Up to know, we've only seen it happen on specific servers, running one
of our test suites. The same test suite on my personal cluster does not
exhibit this issue, whereas the test engineer that encountered this
problem sees it regularly. My servers have SAS drives whereas his have
SATA drives.
On 09/22/2015 08:08 AM, Guido Jäkel wrote:
* Do you use lxc.network.veth.pair to name the hosts side of the veth?
Yes. I rename the veth interfaces to match the names of the containers.
* Was the Container up and running "just before" and you (re)start it within
less than 5min?
Yes. When
On Wed, Sep 23, 2015 at 8:24 AM, Peter Steele wrote:
> On 09/22/2015 08:08 AM, Guido Jäkel wrote:
>>
>> * Do you use lxc.network.veth.pair to name the hosts side of the veth?
>
> Yes. I rename the veth interfaces to match the names of the containers.
>>
>> * Was the
Dear Peter,
* Do you use lxc.network.veth.pair to name the hosts side of the veth?
* Was the Container up and running "just before" and you (re)start it within
less than 5min?
greetings
Guido
___
lxc-users mailing list
On 09/21/2015 03:27 PM, Fajar A. Nugraha wrote:
I remembered something similar a while ago, in ubuntu precise host and
containers, with both lxc 1.0.x and lxc-1.x from ppa. At that time a
container's interface would mysteriously dissapear, including it's
host side veth pair. Only on one
On Tue, Sep 22, 2015 at 5:17 AM, Peter Steele wrote:
> On 09/21/2015 01:20 PM, Peter Steele wrote:
>
>> On 09/21/2015 08:32 AM, Serge Hallyn wrote:
>>
>>> In these cases does /sys/class/net/eth0 exist?
>>>
>>> I'll try to reproduce the condition and check this...
>>
>> I just
On 09/21/2015 01:20 PM, Peter Steele wrote:
On 09/21/2015 08:32 AM, Serge Hallyn wrote:
In these cases does /sys/class/net/eth0 exist?
I'll try to reproduce the condition and check this...
I just checked this. This directory does not exist. There is only an
entry for lo.
Quoting Peter Steele (pwste...@gmail.com):
> We sometimes hit an error where eth0 in a container does not come
> up, leaving the container with only the "lo" device. The system
> messages in the container lists the error
>
> Sep 17 15:58:50 vm-00 network: Bringing up interface eth0: ERROR :
>
On 09/21/2015 08:32 AM, Serge Hallyn wrote:
In these cases does /sys/class/net/eth0 exist?
I'll try to reproduce the condition and check this...
___
lxc-users mailing list
lxc-users@lists.linuxcontainers.org
We sometimes hit an error where eth0 in a container does not come up,
leaving the container with only the "lo" device. The system messages in
the container lists the error
Sep 17 15:58:50 vm-00 network: Bringing up interface eth0: ERROR :
[/etc/sysconfig/network-scripts/ifup-eth] Device eth0
11 matches
Mail list logo