Ok, let’s first try to fix the inter-VM communication, then let’s look at the external network.

Question: 
1) What kind of segmentation are you attempting? VLAN?
     I would recommend you to try VXLAN or GRE first.
     
     In such case you won’t need to modify your local eth0.xx, just specify
the local_ip on the ovs agent configuration, and of course all the other VXLAN bits (that’s 
generally done automatically by packsack)


For the external network

you could plug your eth1 directly to the br-ex bridge, (without vlan tagged), and then specify:

http://docs.openstack.org/havana/install-guide/install/apt/content/install-neutron.configure-networks.plug-in-specific.ovs.vlan.html
--provider:network_type vlan --provider:physical_network physnet1 --provider:segmentation_id SEG_ID
Make sure “physnet1” is the mapping that connects to your br-ex (bridge mappings on the packstack config, that’s
installed to the ovs agent configuration file).

This way neutron would handle the vlan tagging/untagging from/to your external network.

Alternatively, if you want to provide your network node/controller with an external IP address you can try to plug in eth1:1 instead
and keep eth1 for the system (warning, I haven’t tried this specifically, but I guess it may work).

Best,
Miguel Ángel Ajo


On 14/4/2015, at 11:48, pauline phaure <phaurep@gmail.com> wrote:

i turned down the br-ex and br-int and returned the eth0, eth0.xx, eth1,eth1.xx to their original state. Now i can ssh to my servers. But when I spawn a VM it can't reach the dhcp agent.

Starting acpid: OK
cirros-ds 'local' up at 0.84
no results found for mode=local. up 0.87. searched: nocloud configdrive ec2
Starting network...
udhcpc (v1.20.1) started
Sending discover...
Sending discover...
Sending discover...
Usage: /sbin/cirros-dhcpc <up|down>
No lease, failing
WARN: /etc/rc3.d/S40-network failed
cirros-ds 'net' up at 181.06
checking http://169.254.169.254/2009-04-04/instance-id
failed 1/20: up 181.07. request failed
failed 2/20: up 183.09. request failed
failed 3/20: up 185.10. request failed
failed 4/20: up 187.10. request failed
failed 5/20: up 189.11. request failed
failed 6/20: up 191.11. request failed
failed 7/20: up 193.12. request failed
failed 8/20: up 195.12. request failed
failed 9/20: up 197.13. request failed
failed 10/20: up 199.13. request failed
failed 11/20: up 201.14. request failed
failed 12/20: up 203.14. request failed
failed 13/20: up 205.15. request failed
failed 14/20: up 207.15. request failed
failed 15/20: up 209.16. request failed
failed 16/20: up 211.16. request failed
failed 17/20: up 213.17. request failed
failed 18/20: up 215.17. request failed
failed 19/20: up 217.18. request failed
failed 20/20: up 219.18. request failed
failed to read iid from metadata. tried 20
no results found for mode=net. up 221.19. searched: nocloud configdrive ec2
failed to get instance-id of datasource


2015-04-14 10:23 GMT+02:00 pauline phaure <phaurep@gmail.com>:
no actually, packstack just added two files which were empty. Besides my VMs couldn't reach the dhcp agent and it was clear that all the br-ex br-int and br-tun were down.

In order to fill my ifcfg-br-ex file i followed this tuto here https://www.rdoproject.org/Neutron_with_existing_external_network. for the br-int i followed this one https://www.rdoproject.org/forum/discussion/196/quantum-basic-setup/p1. I did exactly the same but i didn't work for me. I think that the problem is related to my interfaces eth0 and eth1 placed in vlans on the physical switch. any idea on how i could fix things?

2015-04-14 10:04 GMT+02:00 Miguel Angel Ajo Pelayo <mangelajo@redhat.com>:
It's been a while since I'm not doing it, but isn't packstack supposed to do that for you?.

What guide steps are you following to modify such files?

----- Original Message -----
>
>
>
> ok, i will but how should i configure this ifcfg-files? should i put the
> br-ex and br-int in the same vlans as eth0 and eth1?
>
> 2015-04-14 10:00 GMT+02:00 Miguel Angel Ajo Pelayo < mangelajo@redhat.com > :
>
>
> Hi Pauline,
>
> I’m afraid that at this point you may need to connect via a KVM or direct
> monitor / keyboard to
> properly reconfigure the ifcfg files.
>
>
> > On 14/4/2015, at 9:48, pauline phaure < phaurep@gmail.com > wrote:
> >
> > Hi everyone,
> > I recently installed Openstack with RDO packstack on two servers. on each
> > server I have 2 interfaces eth0 and eth1 each one of this interfaces is on
> > a seperate vlans. As the VMs spawened by NOVA couldn't get an IP address,
> > I saw in a tutorial that I should edit the files ifcfg-br-ex and
> > ifcfg-br-int and when I did I lost my connection and couldn't anymore ssh
> > to my servers.
> > Do you have any idea how i can solve this?
> > thank you in advance,
> > Pauline,
> > _______________________________________________
> > Rdo-list mailing list
> > Rdo-list@redhat.com
> > https://www.redhat.com/mailman/listinfo/rdo-list
> >
> > To unsubscribe: rdo-list-unsubscribe@redhat.com
>
> Miguel Angel Ajo
>
>
>
>
>
>
> _______________________________________________
> Rdo-list mailing list
> Rdo-list@redhat.com
> https://www.redhat.com/mailman/listinfo/rdo-list
>
> To unsubscribe: rdo-list-unsubscribe@redhat.com



Miguel Angel Ajo