Opening downstream based on a bug from upstream, ASCII art might be messed up in buzgilla but it should be readable elsewhere: Joshua 2017-09-12 12:07:38 UTC Our current deployment has three controllers, each with an odl instance and a neutron/n-odl instance. The neutron/n-odl instances connect to odl via a VIP. In the deployment we have all webwsocket connections are established against odl0 but the neutron port events that trigger the flow creation and set the operational port status to ACTIVE occur on odl2 which is the "leader". This means that the websocket event is generated on odl2 only...but no one is listening for websocket events on odl2...so the port never gets set to ACTIVE in neutron and the VM remains in BUILD state forever and ever. +-----------------------------+ +-----------------------------+ +-----------------------------+ | | | | | | | | | | | | | +---------------------+ | | +---------------------+ | | +---------------------+ | | | | | | | | | | | | | | | odl0 | | | | odl1 | | | | odl2 | | | | | | | | | | | | | | | | | | | | | | | | port created here | | | | | | | | | | | | | | | +----------+----------+ | | +---------------------+ | | +---------------------+ | | ^ | | | | | | | | | | | | | websocket----------------------------+-------------------------------------+ | | | | | | | | | | | | | | | | | | | | +----------+----------+ | | +----------+----------+ | | +-----------+---------+ | | | | | | | | | | | | | | | | | | | | | | | | | | | neutron/n-odl | | | | neutron/n-odl | | | | neutron/n-odl | | | | | | | | | | | | | | | | | | | | | | | | | | | +---------------------+ | | +---------------------+ | | +---------------------+ | | | | | | | | | | | | | +-----------------------------+ +-----------------------------+ +-----------------------------+
Version in which this was observed: opendaylight-6.1.0-2.el7ost
*** Bug 1486917 has been marked as a duplicate of this bug. ***
As a workaround for now, I have proposed patches disabling port status in OOO for HA ODL deployments: https://review.openstack.org/#/c/505788/
Josh, can you please update this BZ? Is their an upstream NetVirt fix?
Is the root of the issue that fact that the ODL "leader" selection is independent of pacemakers selection for HAProxy master?
https://git.opendaylight.org/gerrit/#/c/64602/ Janki, please add the director patches for this. The ones that set the IP of each ODL in xml config.
Tested on an HA deployment. Created multiple instances across multiple networks. The instances correctly switched to ACTIVE, connectivity worked as expected.
(In reply to Josh Hershberg from comment #9) > https://git.opendaylight.org/gerrit/#/c/64602/ > > Janki, please add the director patches for this. The ones that set the IP of > each ODL in xml config. Setting of IP in xml file is done by Puppet-ODL. The patch (Gerrit id 64775) for it is already merged and backported to Carbon. Another patch (Gerrit id 66777) is also pushed to reflect the new file changes and backported to Nitrogen.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2018:2086