Description of problem: Installation failed with byo/config.yml Version-Release number of the following components: [fedora@ip-172-31-33-174 openshift-ansible]$ git log --oneline -1 e4eabd6c9 (HEAD -> release-3.6, tag: openshift-ansible-3.6.173.0.90-1, origin/release-3.6) Automatic commit of package [openshift-ansible] release [3.6.173.0.90-1]. [fedora@ip-172-31-33-174 openshift-ansible]$ ansible --version ansible 2.4.1.0 How reproducible: ansible-playbook -i /tmp/2.file openshift-ansible/playbooks/byo/config.yml Steps to Reproduce: 1. 2. 3. Actual results: Failure summary: 1. Hosts: ec2-34-209-89-50.us-west-2.compute.amazonaws.com, ec2-34-212-31-180.us-west-2.compute.amazonaws.com, ec2-54-200-132-2.us-west-2.compute.amazonaws.com, ec2-54-218-0-139.us-west-2.compute.amazonaws.com Play: Verify Requirements Task: openshift_health_check Message: One or more checks failed Details: check "package_version": Some required package(s) are available at a version that is higher than requested openvswitch-2.8.0 This will prevent installing the version you requested. Please check your enabled repositories or adjust openshift_release. Expected results: Additional info: Please attach logs from ansible-playbook with the -vvv flag
This should be fixed long ago in the release-3.6 branch; https://github.com/openshift/openshift-ansible/pull/5109 was merged 2017-08-17. However, looking a little closer, it looks like that fix didn't address package_version which is what's failing here. I think we may need to also backport https://github.com/openshift/openshift-ansible/pull/6023
https://github.com/openshift/openshift-ansible/pull/6753 merged
When 2.6.1 (10.git20161206.el7fdb) of ovs is installed, it is verified the playbook works now. This fixed the original problem reported in this bz. However, when the following 2.8.0 version is installed, the playbook failed. root@ip-172-31-2-46: ~ # yum list installed | grep openvswitch openvswitch.x86_64 1:2.8.0-4.el7fdb @rhel7-fast-datapath-htb root@ip-172-31-2-46: ~ # yum list installed | grep openshift atomic-openshift.x86_64 3.6.173.0.96-1.git.0.8f6ff22.el7 atomic-openshift-clients.x86_64 atomic-openshift-docker-excluder.noarch atomic-openshift-excluder.noarch atomic-openshift-master.x86_64 atomic-openshift-node.x86_64 3.6.173.0.96-1.git.0.8f6ff22.el7 atomic-openshift-sdn-ovs.x86_64 openshift-ansible.noarch 3.6.99-1.git.0.42f2439.el7 openshift-ansible-callback-plugins.noarch openshift-ansible-docs.noarch openshift-ansible-filter-plugins.noarch openshift-ansible-lookup-plugins.noarch openshift-ansible-playbooks.noarch openshift-ansible-roles.noarch tuned-profiles-atomic-openshift-node.x86_64 The ansible log and journal log on master will be attached.
It looks like OCP 3.6 nodes won't start successfully with ovs 2.8. If that's not going to get fixed by the time ovs 2.8 is out then the check can be updated to inform that it's not a supported combination.
Verified with version openshift-ansible-3.6.173.0.101-1.git.0.7c6b77f.el7, code merged, health check could pass OVS-2.8 installation. By the way, OVS-2.8 still has issue BZ #1527156 cause playbook failed. # rpm -q openvswitch openvswitch-2.8.0-4.el7fdb.x86_64
(In reply to Wenkai Shi from comment #8) > Verified with version openshift-ansible-3.6.173.0.101-1.git.0.7c6b77f.el7, > code merged, health check could pass OVS-2.8 installation. > > By the way, OVS-2.8 still has issue BZ #1527156 cause playbook failed. > > # rpm -q openvswitch > openvswitch-2.8.0-4.el7fdb.x86_64 Thanks for verifying too. Watched the related BZ #1527156 too. Seem likely that the reported issue in Comment 5 is its duplicate.
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-2018:1106
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days