Description of problem: Ansible installer tries to enable etcd_container service even though containerized=false Version-Release number of selected component (if applicable): atomic-openshift-3.1.1.6-3.git.16.5327e56.el7aos.x86_64 How reproducible: Steps to Reproduce: 1. Set OSE v3 installation ansible with containerized=false and etcd hosts. (e.g.) containerized=false ... [OSEv3:children] etcd ... [etcd] ... xxxxx.example.com yyyyy.example.com 2. Run ansible installer 3. Hit the error below TASK: [etcd | Enable etcd] **************************************************** failed: [xxxx.example.com] => {"failed": true} msg: Error when trying to enable etcd_container: rc=1 Failed to execute operation: Access denied Actual results: Hit following error: TASK: [etcd | Enable etcd] **************************************************** failed: [xxxx.example.com] => {"failed": true} msg: Error when trying to enable etcd_container: rc=1 Failed to execute operation: Access denied Expected results: Install without any error Additional info: I will upload users log files in private.
If I executed "systemctl enable etcd_container" instead of "etcd", I got same error "Access denied"[1]. From this, I guess it is a bug of Ansible installer, it should run systemctl enable etcd. If this is a bug, could you please provide us with workaround? [1] # systemctl enable etcd_container Failed to execute operation: Access denied
Sorry, I forgot to tell the ansible version $ cat ansible-rpm-qa-output.txt |grep ansible ansible-1.9.4-1.el7aos.noarch openshift-ansible-roles-3.0.35-1.git.0.6a386dd.el7aos.noarch openshift-ansible-3.0.35-1.git.0.6a386dd.el7aos.noarch openshift-ansible-lookup-plugins-3.0.35-1.git.0.6a386dd.el7aos.noarch openshift-ansible-playbooks-3.0.35-1.git.0.6a386dd.el7aos.noarch openshift-ansible-filter-plugins-3.0.35-1.git.0.6a386dd.el7aos.noarch
Does this happen on openshift-ansible-3.0.47-1.git.0.4498ce3.el7aos ? That's the currently released version.
Yes, it does. The customer confirmed it failed with current released version.
OK, I think I understand this bug. If we set containerized=false in /etc/ansible/hosts as below, some parts of the ansible tries to install by containerized mode. ~~~ containerized=false ~~~ The doc[1] is wrong or installer's bug? https://docs.openshift.com/enterprise/3.1/install_config/install/advanced_install.html#configuring-ansible "If set to false or unset, the default RPM method is used." NOTE: I know there is a workaround to unset containerized=false.
We're discussing https://github.com/openshift/openshift-ansible/pull/1533 as a potential fix for this.
Although I commented on https://github.com/openshift/openshift-ansible/pull/1533, you may hit another containerized installation issue. ~~~ TASK: [openshift_master_ca | Create the master certificates if they do not already exist] *** failed: [master.example.com] => {"cmd": "/usr/local/bin/oadm create-master-certs --hostnames=kubernetes.default,10.67.112.120,kubernetes.default.svc.cluster.local,kubernetes,openshift.default....} msg: [Errno 2] No such file or directory ~~~ I will attach the entire log in private.
(In reply to Kenjiro Nakayama from comment #10) As for the new issue, PR has been opened. https://github.com/openshift/openshift-ansible/pull/1537
Check on openshift-ansible-3.0.57-1.git.0.c633ce7.el7.noarch Install ha-master env with "containerized=false" successfully, move this issue to VERIFIED.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2016:1065