Hi Marius,
is that trick in [2] also working for the discovery image? I’d love to inspect failed
discovery, can I use the same virt-customize on ironic-python-agent.vmlinuz image you
think?
thanks
On 30 Oct 2015, at 13:20, Marius Cornea <marius(a)remote-lab.net>
wrote:
Some good Heat templates debugging tips can be found at:
http://hardysteven.blogspot.co.uk/2015/04/debugging-tripleo-heat-template...
If you're not able to ssh to the controller I'd check first if the
nova instances end up in active state, watch if at any point you're
able to reach them via the ctlplane ip address, maybe the templates
changes break the connectivity. If so you could enable root login via
a firstboot config[1] and access the nodes via console to see what
happened or do this in the image with virt-customize[2]
[1]
https://repos.fedorapeople.org/repos/openstack-m/rdo-manager-docs/liberty...
[2] virt-customize -a overcloud-full.qcow2 --root-password password:rootpass
On Fri, Oct 30, 2015 at 12:39 PM, Ramkumar GOWRISHANKAR
<ramkumar.gowrishankar(a)nuagenetworks.net> wrote:
> Marius,
>
> Ok. I will try that out again. I reduced to just controller since the
> controller + compute was failing for me. I am making changes to the
> templates and images so this is not a stock deployment. But my question
> still stands for debugging as a previous try with controller and fails
> where the nodes are up, the deployment fails in the
> ControllerNodesPostDeployment step and I am not able to ssh into the
> controller and the heat-engine.log is not giving any useful information.
>
> Ramkumar
>
> On Fri, Oct 30, 2015 at 4:12 AM, Marius Cornea <marius(a)remote-lab.net>
> wrote:
>>
>> Hi,
>>
>> You should try at least 1 controller + 1 compute, I don't think a
>> deployment with just one controller is supposed to work.
>>
>> Thanks,
>> Marius
>>
>> On Thu, Oct 29, 2015 at 5:27 PM, Ramkumar GOWRISHANKAR
>> <ramkumar.gowrishankar(a)nuagenetworks.net> wrote:
>>> Hi,
>>>
>>> My virtual test bed deployment with just one controller and no computes
>>> is
>>> failing at ControllerNodesPostDeployment. The debug steps when a
>>> deployment
>>> fails tells to run the following command: "heat resource-show overcloud
>>> ControllerNodesPostDeployment". When I run the command, I see 3 URL
>>> starting
>>> with
http://192.0.2.1:8004.
>>> How do I access these URLs? When I try a wget on these URLs or when I
>>> create
>>> a ssh tunnel from the base machine and try to access the URLs I get
>>> permission denied message. When I try to access just the base URL
>>> (
http://192.0.2.1:8004 mapped to
http://localhost:8005) via a tunnel, I
>>> get
>>> the following message:
>>> {"versions": [{"status":"CURRENT",
"id": "v1.0", "links":
>>>
[{"href":"http://localhost:8005/v1/","rel":"self"}]}]}
>>>
>>> I have looked through the /var/log/heat/ folder for any error messages
>>> but I
>>> cannot find any more detailed error message other than deployment failed
>>> at
>>> step 1 LoadBalancer.
>>>
>>> Any pointers on how to debug a deployment?
>>>
>>> Thanks,
>>>
>>> Ramkumar
>>>
>>> _______________________________________________
>>> Rdo-list mailing list
>>> Rdo-list(a)redhat.com
>>>
https://www.redhat.com/mailman/listinfo/rdo-list
>>>
>>> To unsubscribe: rdo-list-unsubscribe(a)redhat.com
>
>
_______________________________________________
Rdo-list mailing list
Rdo-list(a)redhat.com
https://www.redhat.com/mailman/listinfo/rdo-list
To unsubscribe: rdo-list-unsubscribe(a)redhat.com