Open Stack

I ran into this bug quite a few months ago, too, but worked around it by
loading vhost_net kernel driver.

Currently I get network outages for just a few seconds, like freezes for ~
10-15 seconds, and then everything
works like nothing have ever happened.
I unfortunately can't find anything in the logs inside VM, hypervisor nor
nova-network node.

Regards,
Christian.

On Fri, Oct 5, 2012 at 4:30 PM, Alejandro Comisario <
alejandro.comisario at mercadolibre.com> wrote:

> Hi Cris, maybe your problem is related to this bug ?
>
> https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/997978
>
> Regards.
> Ale
>
>
> On Fri, Oct 5, 2012 at 8:44 AM, Christian Parpart <trapni at gmail.com>wrote:
>
>> Hey all,
>>
>> we're pretty happy about our new OpenStack Essex installation atop of
>> Ubuntu 12.04 (hypervisor and guests).
>> We use KVM (tenant is in a VLAN) as virtualization technology, having
>> about 15 compute nodes, and a
>> central nova-network node to act as the gateway (yet to be fully HA'd,
>> however). On that gateway we're having
>> a PPTP VPN so we can log in on any host or guest though this VPN.
>>
>> Our problem however is, that from time to time (I think it's daily and
>> even multiple times per day)
>> we're encountering kind of networking freezes.
>>
>> I first noticed it as my SSH session froze for a few seconds from my
>> desktop -> VPN (nova-network node) -> KVM guest.
>> I quickly checked others, and they hang, too.
>> I checked a hypervisor, which didn't hang, so it is not a general
>> networking issue (like PPTP or SSH or alike).
>>
>> Now, it feels like there is some problem with networking from and/or to
>> KVM guests and I absolutely have no clue
>> on how to trace this down. It really feels like a bug, but that's really
>> out of my scope, and that's why I'm seeking for advice here,
>> since we just can't stay in this situation :-)
>>
>> It is confirmed that we're having this from desktop -> VPN -> KVM and
>> from physical node (in data center) to KVM.
>> However, I do not yet know whether all KVMs are affected at once (which
>> would indicate that the issue MAY be caused
>> by the central nova-network node) or whether it is just hypervisor based
>> (so it may be due to some hypervisor's state)
>> or just plain random across all 15 compute nodes we have.
>>
>> I now think that this issue may indeed be KVM networking related
>>
>> I'll be happy about any hints and proposals you can provide me in order
>> to track this
>> issue down.
>>
>> Please tell me about any further information you need.
>>
>> Many thanks in advance,
>> Christian Parpart.
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~openstack
>> Post to     : openstack at lists.launchpad.net
>> Unsubscribe : https://launchpad.net/~openstack
>> More help   : https://help.launchpad.net/ListHelp
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20121006/62989f65/attachment.html>