In my case, it was due to the messaging protocol mismatch. As soon as I
ensured that the API server recognized all agents at the latest level, the
problem cleared. The symptoms seem similar, so I expect even though this is
a new installation, you may still be hitting this somehow. For example...
have you pinned the messaging version in the conf files?
For me, it was a lot of hours of tracking down exactly what was going on. I
eventually traced it into the database where I found rows that recorded the
nodes at older messaging versions, and they were machines I didn't even
have online as they had been replaced long ago. Clearing those out was all
that it took...
On Mon, Aug 7, 2017 at 11:12 AM, Sven Michels <email address hidden> wrote:
> Hey Mark,
>
> but we're having a new installation... so this should not be the
> case in our case. So you found nothing else related to it?
>
> Cheers and thanks,
> Sven
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1682423
>
> Title:
> create instance with dashbaord / 404 returned
>
> Status in OpenStack Compute (nova):
> Incomplete
>
> Bug description:
> Description
> ===========
> When launching an instance over dashboard, a 404 code is returned (
> Instance XX-XX-XX could not be found ).
> Same issue occurs when you make it through nova client, but it has no
> effect.
> In both tasks the instance is created.
> You can see a "HTTP exception thrown:" in the nova-api logfile.
>
>
> Steps to reproduce
> ==================
> A chronological list of steps which will bring off the
> issue you noticed:
> * upgraded all nodes to CentoS 7.3
> * no issues
> * upgraded from newton to ocata (RDO)
> * 404 issue
>
>
> * launching an instance through dashboard.
> The instance has been added but a 404 returned.
> So an error is displayed in dashboard interface (see
> create_instance_error.png)
>
> For testing purposes i used curl.
> see output.txt
>
>
> * launching an instance using nova client
> Same error occurs but it has no effect.
>
> see output.txt
>
>
> Expected result
> ===============
> A new instance launched and the wizard should be closed.
>
> Actual result
> =============
> I get an 404 Code and the "instance wizard" is still open although the
> instances was created and is running.
>
> Environment
> ===========
> openstack-nova-console-15.0.0-1.el7
> python-nova-15.0.0-1.el7
> openstack-nova-placement-api-15.0.0-1.el7
> python2-novaclient-7.1.0-1.el7
> openstack-nova-scheduler-15.0.0-1.el7
> openstack-nova-conductor-15.0.0-1.el7
> openstack-nova-novncproxy-15.0.0-1.el7
> openstack-nova-api-15.0.0-1.el7
> openstack-nova-common-15.0.0-1.el7
> openstack-nova-cert-15.0.0-1.el7
> openstack-dashboard-11.0.0-1.el7
> python-networking-bgpvpn-dashboard-6.0.0-0.1.681a5e7git.el7
>
>
> Logs & Configs
> ==============
> see nova-api-logs.txt
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/nova/+bug/1682423/+subscriptions
>
In my case, it was due to the messaging protocol mismatch. As soon as I
ensured that the API server recognized all agents at the latest level, the
problem cleared. The symptoms seem similar, so I expect even though this is
a new installation, you may still be hitting this somehow. For example...
have you pinned the messaging version in the conf files?
For me, it was a lot of hours of tracking down exactly what was going on. I
eventually traced it into the database where I found rows that recorded the
nodes at older messaging versions, and they were machines I didn't even
have online as they had been replaced long ago. Clearing those out was all
that it took...
On Mon, Aug 7, 2017 at 11:12 AM, Sven Michels <email address hidden> wrote:
> Hey Mark, /bugs.launchpad .net/bugs/ 1682423 instance_ error.png) nova-console- 15.0.0- 1.el7 nova-15. 0.0-1.el7 nova-placement- api-15. 0.0-1.el7 novaclient- 7.1.0-1. el7 nova-scheduler- 15.0.0- 1.el7 nova-conductor- 15.0.0- 1.el7 nova-novncproxy -15.0.0- 1.el7 nova-api- 15.0.0- 1.el7 nova-common- 15.0.0- 1.el7 nova-cert- 15.0.0- 1.el7 dashboard- 11.0.0- 1.el7 networking- bgpvpn- dashboard- 6.0.0-0. 1.681a5e7git. el7 /bugs.launchpad .net/nova/ +bug/1682423/ +subscriptions
>
> but we're having a new installation... so this should not be the
> case in our case. So you found nothing else related to it?
>
> Cheers and thanks,
> Sven
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https:/
>
> Title:
> create instance with dashbaord / 404 returned
>
> Status in OpenStack Compute (nova):
> Incomplete
>
> Bug description:
> Description
> ===========
> When launching an instance over dashboard, a 404 code is returned (
> Instance XX-XX-XX could not be found ).
> Same issue occurs when you make it through nova client, but it has no
> effect.
> In both tasks the instance is created.
> You can see a "HTTP exception thrown:" in the nova-api logfile.
>
>
> Steps to reproduce
> ==================
> A chronological list of steps which will bring off the
> issue you noticed:
> * upgraded all nodes to CentoS 7.3
> * no issues
> * upgraded from newton to ocata (RDO)
> * 404 issue
>
>
> * launching an instance through dashboard.
> The instance has been added but a 404 returned.
> So an error is displayed in dashboard interface (see
> create_
>
> For testing purposes i used curl.
> see output.txt
>
>
> * launching an instance using nova client
> Same error occurs but it has no effect.
>
> see output.txt
>
>
> Expected result
> ===============
> A new instance launched and the wizard should be closed.
>
> Actual result
> =============
> I get an 404 Code and the "instance wizard" is still open although the
> instances was created and is running.
>
> Environment
> ===========
> openstack-
> python-
> openstack-
> python2-
> openstack-
> openstack-
> openstack-
> openstack-
> openstack-
> openstack-
> openstack-
> python-
>
>
> Logs & Configs
> ==============
> see nova-api-logs.txt
>
> To manage notifications about this bug go to:
> https:/
>
--
Mark Mielke <email address hidden>