Arash,
Please, create BZ for this issue. I don't have that environment in meantime.
Thanks.
Boris.
Date: Wed, 29 Apr 2015 17:26:24 +0200
From: kchamart(a)redhat.com
To: bderzhavets(a)hotmail.com
CC: ak(a)cloudssky.com; apevec(a)gmail.com; rdo-list(a)redhat.com
Subject: Re: [Rdo-list] RE(3): RE(2) : RDO Kilo RC snapshot - core packages
On Sun, Apr 26, 2015 at 12:52:36PM -0400, Boris Derzhavets wrote:
> Confirmed version as of 04/21/2015 ( via delorean repo ) was tested
> with answer-file for Two Node install Controller&&Network and Compute
> Answer-file sample was brought from blog entry ( works fine on Juno)
>
http://bderzhavets.blogspot.com/2015/04/switching-to-ethx-interfaces-on-f...
>
> Packstack completed OK . However, table "compute_nodes" in nova
> database appeared to be empty up on completion. System doesn't see
> Hypervisor Host either via Dashboard or via nova CLI as mentioned by
> Arash. So it cannot scheduler instance, no matter that `nova
> service-list` is OK and openstack-nova-compute is up and running on
> Compute Node. I've also I've got same "WARNING
> nova.compute.resource_tracker [-] No service record for host compute1"
> in /var/log/nova/nova-compute.log on Compute node.
Boris, if you're able to reproduce this consistently, can you please
file a Nova bug with clear reproducer details? Along with contextual log
files w/ debug enabled.
--
/kashyap