[Openstack-operators] I have finally created an instance, and it works! However, there is no ethernet card
Andreas Scheuring
scheuran at linux.vnet.ibm.com
Mon Aug 18 07:04:26 UTC 2014
Assuming you're running a libvirt based hypervisor (e.g. kvm). Could you
please dump the libvirt xml of your instance?
You can get it this way
virsh list --all
--> shows a list of all virtual servers running on your hypervisor
virsh dumpxml <vm-id>
--> dumps the xml of your vm, addressed by the id (first column). Id
does not correlate with UUID!!!!! So if you're not sure which list entry
belongs to your instance just stop all other via openstack to have only
one running.
PS: on some systems you have to sudo when using virsh.
There should be a subtag of devices called <interface type='bridge'> or
'network' or somehting like that representing your eth interface
Andreas
On Fri, 2014-08-15 at 11:43 -0700, Jeff Silverman wrote:
> I have been surfing the internet, and one of the ideas that comes to
> mind is modifying the /etc/neutron/agent.ini file on the compute
> nodes. In the agent.ini file, there is a comment near the top that is
> almost helpful:
>
>
> # L3 requires that an interface driver be set. Choose the one that
> best
> # matches your plugin.
>
>
> The only plug I know about is ml2. I have no idea if that is right
> for me or not. And I have no idea to choose the interface drive that
> best matches my plugin.
>
>
> Thank you
>
>
>
>
> Jeff
>
>
>
>
>
>
>
>
>
>
> On Fri, Aug 15, 2014 at 10:26 AM, Jeff Silverman <jeff at sweetlabs.com>
> wrote:
> By "defined a network space for your instances", does that
> mean going through the process as described
> in http://docs.openstack.org/icehouse/install-guide/install/yum/content/neutron-ml2-compute-node.html?
>
>
> I got part way through that when I realized that the procedure
> was going to bridge packets through neutron. That's not what
> I want. I want the packets to go directly to the physical
> router. For example, I have two tenants, with IP addresses
> 10.50.15.80/24 and 10.50.18.15.90/24.and the router is at
> 10.50.15.1. There is a nice picture of what I am trying to do
> at http://docs.openstack.org/openstack-ops/content/network_troubleshooting.html#nova_network_traffic_in_cloud . But if the hypervisor doesn't present a virtual device to the guests, then nothing else is going to happen. The network troubleshooting guide http://docs.openstack.org/openstack-ops/content/network_troubleshooting.html#nova_network_traffic_in_cloud does not explain what to do if the virtual NIC is missing.
>
>
>
>
> Thank you
>
>
> Jeff
>
>
>
>
> On Fri, Aug 15, 2014 at 9:38 AM, Abel Lopez
> <alopgeek at gmail.com> wrote:
> Curious if you’ve defined a network space for your
> instances. If you’re using the traditional
> flat_network, this is known as the ‘fixed_address’
> space.
> If you’re using neutron, you would need to create a
> network and a subnet (and router with gateway, etc).
> You’d then assign the instance to a network at launch
> time.
>
>
>
> On Aug 15, 2014, at 9:17 AM, Jeff Silverman
> <jeff at sweetlabs.com> wrote:
>
> > <ip_a.png>
> >
> >
> >
> > For those of you that can't see pictures:
> > $ sudo ip a
> > 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc
> > noqueue
> > link/loopback 00:00:00:00:00:00 brd
> > 00:00:00:00:00:00
> > inet 127.0.0.1/8 scope host lo
> > inet6 ::1/128 scope host
> > valid_lft forever preferred_1ft forever
> >
> >
> > I suspect that the issue is that the hypervisor is
> > not presenting a virtual ethernet card.
> >
> >
> > Thank you
> >
> >
> >
> >
> > Jeff
> >
> >
> >
> >
> > On Thu, Aug 14, 2014 at 6:57 PM, Nhan Cao
> > <nhanct92 at gmail.com> wrote:
> > can you show output of command:
> > ip a
> >
> >
> >
> >
> >
> >
> > 2014-08-15 7:41 GMT+07:00 Jeff Silverman
> > <jeff at sweetlabs.com>:
> > People,
> >
> >
> > I have brought up an instance, and I
> > can connect to it using my browser!
> > I am so pleased.
> >
> >
> > However, my instance doesn't have an
> > ethernet device, only a loopback
> > device. My management wants me to
> > use a provider network, which I
> > understand to mean that my instances
> > will have IP addresses in the same
> > space as the controller, block
> > storage, and compute node
> > administrative addresses. However,
> > I think that discussing addressing
> > is premature until I have a working
> > virtual ethernet card.
> >
> >
> > I am reading
> > through http://docs.openstack.org/icehouse/install-guide/install/yum/content/neutron-ml2-compute-node.html and I think that the ML2 plugin is what I need. However, I think I do not want a network type of GRE, because that encapsulates the packets and I don't have anything to un-encapsulate them.
> >
> >
> > Thank you
> >
> >
> >
> >
> > Jeff
> >
> >
> >
> >
> > --
> > Jeff Silverman
> > Systems Engineer
> > (253) 459-2318 (c)
> >
> >
> >
> >
> > _______________________________________________
> > OpenStack-operators mailing list
> > OpenStack-operators at lists.openstack.org
> > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
> >
> >
> >
> >
> >
> >
> >
> > --
> > Jeff Silverman
> > Systems Engineer
> > (253) 459-2318 (c)
> >
> >
> > _______________________________________________
> > OpenStack-operators mailing list
> > OpenStack-operators at lists.openstack.org
> > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
> >
>
>
>
>
>
>
> --
> Jeff Silverman
> Systems Engineer
> (253) 459-2318 (c)
>
>
>
>
>
>
> --
> Jeff Silverman
> Systems Engineer
> (253) 459-2318 (c)
>
>
> _______________________________________________
> OpenStack-operators mailing list
> OpenStack-operators at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
More information about the OpenStack-operators
mailing list