O->P fails during AllNodesPostUpgradeSteps.ControllerDeployment_Step1 on haproxy docker.

[Sofer Athlan-Guyot]

Hi,

during upgrade from O->P we have AllNodesPostUpgradeSteps.ControllerDeployment_Step1 failing[1] with[2]:

           "2017-11-30 18:49:40,042 INFO: 122731 -- Pulling image: 192.168.24.1:8787/pike/centos-binary-haproxy:1e0dd6786e3a4326a37e47541844657524e689a9_81220c5f",
            "2017-11-30 18:49:40,042 INFO: 122732 -- Pulling image: 192.168.24.1:8787/pike/centos-binary-mariadb:1e0dd6786e3a4326a37e47541844657524e689a9_81220c5f",
            "2017-11-30 18:49:40,042 INFO: 122733 -- Pulling image: 192.168.24.1:8787/pike/centos-binary-nova-compute:1e0dd6786e3a4326a37e47541844657524e689a9_81220c5f",
            "2017-11-30 18:50:00,710 ERROR: 122731 -- Failed running docker-puppet.py for haproxy",
            "2017-11-30 18:50:00,710 ERROR: 122731 -- Notice: hiera(): Cannot load backend module_data: cannot load such file -- hiera/backend/module_data_backend",
            "2017-11-30 18:50:00,710 ERROR: 122731 -- + mkdir -p /etc/puppet",
            "+ cp -a /tmp/puppet-etc/auth.conf /tmp/puppet-etc/hiera.yaml /tmp/puppet-etc/hieradata /tmp/puppet-etc/modules /tmp/puppet-etc/puppet.conf /tmp/puppet-etc/ssl /etc/puppet",
            "+ rm -Rf /etc/puppet/ssl",
            "+ echo '{\"step\": 6}'",

[1] https://logs.rdoproject.org/08/523408/6/openstack-check/gate-tripleo-ci-centos-7-containers-multinode-upgrades-pike/Z144e9e1f762c4aeba5d375dca6ef5631/undercloud/home/jenkins/overcloud_upgrade_console.log.txt.gz#_2017-11-30_18_51_30

[2] full log: https://logs.rdoproject.org/08/523408/6/openstack-check/gate-tripleo-ci-centos-7-containers-multinode-upgrades-pike/Z144e9e1f762c4aeba5d375dca6ef5631/undercloud/home/jenkins/failed_upgrade.log.txt.gz

[Sofer Athlan-Guyot (sofer-athlan-guyot)]

Maybe https://review.openstack.org/#/q/I963aede41403ebbe3b9afb55a725b304a30a0cbb would solve this.

[Sofer Athlan-Guyot (sofer-athlan-guyot)]

The root cause is due to puppet-pacemaker not being updated.

We have puppet-pacemaker-0.6.1-0.20171124095828.5f56a96.el7.centos.noarch installed for ocata and we need to upgrade to puppet-pacemaker-0.6.1-0.20171103172230.9a46ecd.el7.centos.noarch.rpm (in pike delorean, there for instance https://trunk.rdoproject.org/centos7-pike/6f/18/6f18d036d2ff6c5831b5e114a252a5f4c67702ea_09a4208c/)

But this is not happening, maybe because 20171124095828 is higher than 20171103172230 and 0.6.1-0 is the same for both ?

5f56a96 is the tip of stable/0.6.x which matches what is defined in https://github.com/redhat-openstack/rdoinfo/blob/master/rdo.yml#L1275-L1286 but the naming is not 0.5.0.

Checking with rdo as what is happening there.

[Sofer Athlan-Guyot (sofer-athlan-guyot)]

Solution would be that https://review.rdoproject.org/r/#/c/11269/ . We get 0.7.0 in the pike package name so we are certain that the package upgrade.

Here the issue is that the ocata branch get updated after the pike commit. As there are both 0.6.0 in metadata.json, the ocata one wins.

[Alex Schultz (alex-schultz)]

Marking fixed as https://review.rdoproject.org/r/#/c/11269/ was merged