[PVE-User] veths/gateway failing after random time.

[dan]

On Tue, Apr 17, 2012 at 8:56 AM, dan <dandenson at gmail.com> wrote:
> nevermind, failed on that machine too.  I will look at my local
> routing and make sure it isn't some unrelated problem...
>
> On Tue, Apr 17, 2012 at 8:46 AM, dan <dandenson at gmail.com> wrote:
>> On Tue, Apr 17, 2012 at 7:55 AM, dan <dandenson at gmail.com> wrote:
>>>> Verify MAC
>>>
>>> I have checked and each container has a different MAC address that
>>> doesn't match anything else on the network.
>>>
>>> On Tue, Apr 17, 2012 at 12:36 AM, Timh B <timh at shiwebs.net> wrote:
>>>> Hi,
>>>>
>>>> We've seen this as well, but only on IPv6 as we get a fair amount of
>>>> traffic over IPv4 - an "inside ping" from the vmnode actually makes it
>>>> work again, we have no idea what is causing this and for us where we are
>>>> really anxious we do a cron-scheduled ping to the "world" every 5 minutes.
>>>> It might have something to do with power-saving stuff as we haven't seen
>>>> this on older ARCH cpu's, only new ones and our current working theory is
>>>> that apm/powermanager "kills" fds/procs that has been idle for longer than
>>>> X.
>>>>
>>>> If it's not the case, then we might have a bug on our hands.
>>>>
>>>> //T
>>>
>>> I can't ping out of the container to the internet as you suggest.  I
>>> can ping the local network just fine, and the local network can also
>>> ping the container.
>>>
>>>
>>> my NIC in the host is a RTL-8169 gigabit nic.
>>
>>
>> I have two identical machines that this is happening on.  I migrated
>> to a 3rd machine that is a different platform and it appears to be
>> fine.  No drops in 30 minutes.  I'll change out the network card and
>> see if that is the issue.  Realtek cards have given me grief in other
>> ways in the past...

FYI, this is a totally different platform.  first 2 boxes are AMD X2
machines with VT on an nforce cpu and the 3rd is a Core2 Duo without
VT.