O->P upgrade: ocata deploy fails because of port 8000 collision

[Jiří Stránský]

http://logs.openstack.org/61/460061/11/experimental/gate-tripleo-ci-centos-7-containers-multinode-upgrades-nv/13a0976/logs

In O->P upgrade job, deploying Ocata failed because of httpd failing to start, because of collision on port 8000.

If i understand correctly, Ocata shouldn't be running Heat under httpd yet, but i do see the config files generated there in /etc/httpd/conf.d:

 [PARENTDIR] Parent Directory -
[ ] 00-nova-placement-api.conf.txt.gz 2017-05-05 12:54 159
[ ] 10-cinder_wsgi.conf.txt.gz 2017-05-05 12:54 573
[ ] 10-heat_api_cfn_wsgi.conf.txt.gz 2017-05-05 12:54 705
[ ] 10-heat_api_cloudwatch_wsgi.conf.txt.gz 2017-05-05 12:54 679
[ ] 10-heat_api_wsgi.conf.txt.gz 2017-05-05 12:54 660
[ ] 10-keystone_wsgi_admin.conf.txt.gz 2017-05-05 12:54 573
[ ] 10-keystone_wsgi_main.conf.txt.gz 2017-05-05 12:54 575
[ ] 10-placement_wsgi.conf.txt.gz 2017-05-05 12:54 587
[ ] 15-default.conf.txt.gz 2017-05-05 12:54 423
[ ] ssl.conf.txt.gz 2017-05-05 12:54 376

I double checked that the overcloud-custom-tht script checks out ocata THT for the initial deployment:

2017-05-05 12:32:30 | + rm -rf /home/jenkins/tripleo-heat-templates
2017-05-05 12:32:30 | + git clone -b stable/ocata --single-branch https://git.openstack.org/openstack/tripleo-heat-templates /home/jenkins/tripleo-heat-templates
2017-05-05 12:32:30 | Cloning into '/home/jenkins/tripleo-heat-templates'...
2017-05-05 12:32:32 | + '[' '!' -e /home/jenkins/tripleo-heat-templates/ci/environments/multinode-container-upgrade.yaml -a -e /usr/share/openstack-tripleo-heat-templates/ci/environments/multinode-container-upgrade.yaml ']'
2017-05-05 12:32:32 | + cp /usr/share/openstack-tripleo-heat-templates/ci/environments/multinode-container-upgrade.yaml /home/jenkins/tripleo-heat-templates/ci/environments/multinode-container-upgrade.yaml

And this directory is indeed being used to deploy:

2017-05-05 12:33:43 | + openstack overcloud deploy --templates /home/jenkins/tripleo-heat-templates --libvirt-type qemu --timeout 80 -e /home/jenkins/cloud-names.yaml -e /home/jenkins/tripleo-heat-templates/environments/deployed-server-environment.yaml -e /home/jenkins/tripleo-heat-templates/environments/deployed-server-bootstrap-environment-centos.yaml -e /home/jenkins/tripleo-heat-templates/ci/environments/multinode-container-upgrade.yaml -e /home/jenkins/tripleo-heat-templates/environments/low-memory-usage.yaml --validation-errors-nonfatal -e /opt/stack/new/tripleo-ci/test-environments/worker-config.yaml -e /usr/share/openstack-tripleo-heat-templates/environments/debug.yaml --compute-scale 0 --overcloud-ssh-user jenkins

[Jiří Stránský (jistr)]

This is addressed in latest https://review.openstack.org/#/c/460061/

[Martin André (mandre)]

Should we close the bug now that https://review.openstack.org/#/c/460061/ has landed?

[Jiří Stránský (jistr)]

Yes, thanks for the reminder. Same with https://bugs.launchpad.net/tripleo/+bug/1685795