multi nodes provider network ovs config [closed]
Hi list,
I have configured a two node clusters (using a provider network) i can see both compute nodes with nova. I had to remove br-ex config because of the associated bug ( –os-neutron-ovs-bridge-interfaces=br-ex:eth0) and completed the network configuration manually.
I can launch instances on the node 1 , but I think i'm missing a ovs config on the second node because when i try to launch a new instance on the second node, i'm getting this error in neutron :
2016-08-09 13:37:16.032 2441 INFO neutron.agent.common.ovs_lib [req-0c6ac8fc-8497-4a5e-9d56-ea522cf650c2 - - - - -] Port d8faee3a-8ddb-4e98-9965-7ed8e556f1ce not present in bridge br-int
2016-08-09 13:37:16.033 2441 INFO neutron.plugins.ml2.drivers.openvswitch.agent.ovs_neutron_agent [req-0c6ac8fc-8497-4a5e-9d56-ea522cf650c2 - - - - -] port_unbound(): net_uuid None not in local_vlan_map
2016-08-09 13:37:16.033 2441 INFO neutron.agent.securitygroups_rpc [req-0c6ac8fc-8497-4a5e-9d56-ea522cf650c2 - - - - -] Remove device filter for [u'd8faee3a-8ddb-4e98-9965-7ed8e556f1ce']
What should i do to fix this issue ?
Thanks
Michael
PLease, post ml2_conf.ini && openvswitch-agent.ini on both nodes, comment which one openvswitch-agent.ini is set up on Controller which one on Compute. Check neutron-ovs-agent status on failing node.
This question is about a version of OpenStack which has been designated End Of Life - https://releases.openstack.org/ If you're still experiencing the problem with a more recent release, please reopen, or make a new ticket with updated details. Thanks!