[Rdo-list] [rdo-manager] Authentication required during overcloud deployment

[Marius Cornea]

----- Original Message -----
> From: "Dan Sneddon" <dsneddon at redhat.com>
> To: rdo-list at redhat.com
> Sent: Thursday, October 15, 2015 1:23:52 AM
> Subject: Re: [Rdo-list] [rdo-manager] Authentication required during overcloud deployment
> 
> On 10/14/2015 03:03 PM, Erming Pei wrote:
> > Hi,
> > 
> >   I am deploying the overcloud in baremetal way and after a couple of
> > hours, it showed:
> > 
> > $ openstack overcloud deploy --templates
> > Deploying templates in the directory
> > /usr/share/openstack-tripleo-heat-templates
> > ^[[A^[[BERROR: openstack ERROR: Authentication failed. Please try again
> > with option --include-password or export HEAT_INCLUDE_PASSWORD=1
> > Authentication required
> > 
> > 
> > But I checked the nodes are now running:
> > 
> > [stack at gcloudcon-3 ~]$ nova list
> > +--------------------------------------+-------------------------+--------+------------+-------------+--------------------+
> > 
> > | ID                                   | Name                    |
> > Status | Task State | Power State | Networks           |
> > +--------------------------------------+-------------------------+--------+------------+-------------+--------------------+
> > 
> > | 1ba04ac0-fe2b-4318-aa31-2e5f4d8422a6 | overcloud-controller-0  |
> > ACTIVE | -          | Running     | ctlplane=10.0.6.60 |
> > | c152ba59-3aed-4fb0-81fa-e3fed7e35cf6 | overcloud-novacompute-0 |
> > ACTIVE | -          | Running     | ctlplane=10.0.6.61 |
> > +--------------------------------------+-------------------------+--------+------------+-------------+--------------------+
> > 
> > 
> > 1. Should I re-deploy the nodes or there is a way to do update/makeup
> > for the authentication issue?
> > 
> > 2.
> > I don't know how to access to the nodes.
> > There is not an overcloudrc file produced.
> > 
> > $ ls overcloud*
> > overcloud-env.json  overcloud-full.initrd  overcloud-full.qcow2
> > overcloud-full.vmlinuz
> > 
> > overcloud-full.d:
> > dib-manifests
> > 
> > Is it via ssh key or password? Should I set the authentication method
> > somewhere?
> > 
> > 
> > 
> > Thanks,
> > 
> > Erming
> > 
> > 
> > _______________________________________________
> > Rdo-list mailing list
> > Rdo-list at redhat.com
> > https://www.redhat.com/mailman/listinfo/rdo-list
> > 
> > To unsubscribe: rdo-list-unsubscribe at redhat.com
> 
> This error generally means that something in the deployment got stuck,
> and the deployment hung until the token expired after 4 hours. When
> that happens, there is no overcloudrc generated (because there is not a
> working overcloud). You won't be able to recover with a stack update,
> you'll need to perform a stack-delete and redeploy once you know what
> went wrong.
> 
> Generally a deployment shouldn't take anywhere near that long, a bare
> metal deployment with 6 hosts takes me less than an hour, and less than
> 2 including a Ceph deployment. In fact, I usually set a timeout using
> the --timeout option, because if it hasn't finished after, say 90
> minutes (depending on how complicated the deployment is), then I want
> it to bomb out so I can diagnose what went wrong and redeploy.
> 
> Often when a deployment times out it is because there were connectivity
> issues between the nodes. Since you can log in to the hosts, you might
> want to do some basic network troubleshooting, such as:
> 
> $ ip address   # check to see that all the interfaces are there, and
> that the IP addresses have been assigned
> 
> $ sudo ovs-vsctl show     # make sure that the bridges have the proper
> interfaces, vlans, and that all the expected bridges show up
> 
> $ ping <other overcloud nodes>  # you can try this on all VLANs to make
> sure that any VLAN trunks are working properly
> 
> $ sudo ovs-appctl bond/show  # if running bonding, check to see the
> bond status
> 
> $ sudo os-net-config --debug -c /etc/os-net-config/config.json   # run
> the network configuration script again to make sure that it is able to
> configure the interfaces without error. WARNING, MAY BE DISRUPTIVE as
> this will reset the network interfaces, run on console if possible.

Also looking for erros in os-collect-config logs might prove to be useful( journalctl -u os-collect-config ), more details on the debugging docs:
https://repos.fedorapeople.org/repos/openstack-m/rdo-manager-docs/liberty/troubleshooting/troubleshooting-overcloud.html

> However, I want to first double-check that you had a valid command
> line. You only show "openstack deploy overcloud --templates" in your
> original email. You did have a full command-line, right? Refer to the
> official installation guide for the right parameters.
> 
> --
> Dan Sneddon         |  Principal OpenStack Engineer
> dsneddon at redhat.com |  redhat.com/openstack
> 650.254.4025        |  dsneddon:irc   @dxs:twitter
> 
> _______________________________________________
> Rdo-list mailing list
> Rdo-list at redhat.com
> https://www.redhat.com/mailman/listinfo/rdo-list
> 
> To unsubscribe: rdo-list-unsubscribe at redhat.com
>