On Wed, Nov 25, 2015 at 11:57 AM, Simone Tiraboschi <stirabos@redhat.com> wrote:



Hello Simone,
always trying to setup a working environment from master using self hosted engine
I'm using the appliance at the link you provided, because it seems that the ovirt-appliance rpm is outside from master repo (I suppose due to its size and daily mirrors effort?)

It's basically for the size, we'll fix it when we'll move to new infrastructure.
 

But it seems it remains stuck at

[ INFO  ] Connecting to the Engine
[ INFO  ] Waiting for the host to become operational in the engine. This may take several minutes...

in this phase I see that the engine vm is not up

This looks strange, the engine should be up at that point.
 
In /var/log/libvirt/qemu/HostedEngine.log
2015-11-24 10:51:15.576+0000: shutting down

Can you please attach vdsm logs and /var/log/messages from your host to let me check why the VM died?
 

and both ovirt-ha-broker and ovirt-ha-agent are down and yet disable at host side. 

and at the end
[ INFO  ] Still waiting for VDSM host to become operational...
...
[ INFO  ] Still waiting for VDSM host to become operational...
[ ERROR ] Timed out while waiting for host to start. Please check the logs.
[ ERROR ] Unable to add hosted_engine_1 to the manager
[ INFO  ] Saving hosted-engine configuration on the shared storage domain
[ INFO  ] Shutting down the engine VM  
[ INFO  ] Enabling and starting HA services
          Hosted Engine successfully set up
[ INFO  ] Stage: Clean up
[ INFO  ] Generating answer file '/var/lib/ovirt-hosted-engine-setup/answers/answers-20151124123248.conf'
[ INFO  ] Generating answer file '/etc/ovirt-hosted-engine/answers.conf'
[ INFO  ] Stage: Pre-termination
[ INFO  ] Stage: Termination

Perhaps is it better to configure a separate self hosted engine os and not use the appliance?
Or all the sef hosted engine in master is broken right now and I have to wait?

I tryied it a couple of days ago with the appliance and it was OK.
I think is worth now to had CI jobs also from master.
 
On the same infra if I configure 3.6.0 in the same way I have no problem.
I'm working at snapshot level in virt-manager so I can revert back just to before the phase where I enable of 3.6-master repo on host.

Let me know if you want me to test anything. 

Gianluca


Actually after I manually started ovirt-ha-broker and ovirt-ha-agent at host side, the engine vm has been started.
And I was able to: access it, put infra in maintenanance, shutdown all and then restart host and exit form maintenance with the engine vm started again and accessible....
I also see italian shown as a possible language, so it is indeed master version and I can begin to crosscheck... ;-)

To analyze the ERRORS I got from host deploy phase, you can find

messages in tar.gz format here:
https://drive.google.com/file/d/0BwoPbcrMv8mvckRNbUdzWDQ2bWc/view?usp=sharing

ovirt-hosted-engine logs in tar.gz format here:
https://drive.google.com/file/d/0BwoPbcrMv8mvZzVBZXBNRzlvRkk/view?usp=sharing

vdsm logs tree in tar.gz format here:
https://drive.google.com/file/d/0BwoPbcrMv8mvcGZjSFFlTDd1aDA/view?usp=sharing

engine side logs in tar.gz format here:
https://drive.google.com/file/d/0BwoPbcrMv8mvTHBqMjlCaktOUzA/view?usp=sharing

Output of some commands (let me know if you want me to run other ones to crosscheck if the installation is ok from all points of view):

[root@ovc71 qemu]# hosted-engine --check-liveliness
Hosted Engine is up!

[root@ovc71 qemu]# hosted-engine --vm-status


--== Host 1 status ==--

Status up-to-date                  : True
Hostname                           : ovc71.localdomain.local
Host ID                            : 1
Engine status                      : {"health": "good", "vm": "up", "detail": "up"}
Score                              : 3400
stopped                            : False
Local maintenance                  : False
crc32                              : 6a8e8808
Host timestamp                     : 1510

due to the bug that the storage domain of hosted engine is not shown inside the web admin portal I cannot crosscheck from the gui....

But from the gui itself the host results as unresponsive (enjoy the italian ... ;-) so there is indeed something not correct, I hope fixable without reinstall... see print screen

https://drive.google.com/file/d/0BwoPbcrMv8mvUEpRS2RteXhiWjQ/view?usp=sharing

Gianluca