Instance unable to receive traffic using linuxbridge

asked 2014-04-10 13:22:44 -0600

hutaim gravatar image

I have a working cloud set up using Grizzly and linuxbridge on SLES. I am setting up another staging environment using RDO Havana on CentOS 6.5 and trying to match much of the configuration from the existing cloud. I can boot instances in this new staging environment and traffic appears to leave the instance fine, but the instance doesn't receive anything.

The compute hosts have only a single NIC and the network is controlled external to OpenStack. I have zero control over the IPs on the network. Using linuxbridge and the Noop firewall means OpenStack has the wrong IPs recorded for instances, but the lack of firewall allows things to keep working. I'm not committed to using linuxbridge over something like OVS, but linuxbridge is what worked with the least effort when upgrading to Grizzly.

Using tshark, I was able to confirm DHCP replies to the instance that never get answered.

edit retag flag offensive close merge delete


please check your ip tables for any drop rule.

SGPJ gravatar imageSGPJ ( 2014-04-10 13:28:47 -0600 )edit

I am using the NoopFirewallDriver and have the iptables service stopped. Plus the security group is wide open, allowing ports 1-65535 for UDP and TCP.

hutaim gravatar imagehutaim ( 2014-04-10 14:14:36 -0600 )edit

Actually, I think the instance does receive traffic. Managed to get the console working and it has the right IP. However, the network is unresponsive despite wide open firewall.

hutaim gravatar imagehutaim ( 2014-04-10 14:52:31 -0600 )edit

1 answer

Sort by ยป oldest newest most voted

answered 2014-05-05 11:27:55 -0600

hutaim gravatar image

updated 2014-05-06 08:36:42 -0600

It was a different tables: (ebtables). Adjusting the rules to prevent spoofing fixed things.

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2014-04-10 13:22:44 -0600

Seen: 171 times

Last updated: May 06 '14