Patch to work with Origin

Make use of AnsibleDumper in openshift_master filters s.t. we can represent

More reliable wait for master after full host reboot. (dgoodwin@redhat.com)
Backport latest openshift-cert-expiry role updates to 1.4
increase test coverage (jdetiber@redhat.com)
add test for utils to bump coverage (jdetiber@redhat.com)
update travis to use tox for utils (jdetiber@redhat.com)
Correct tox to run on Travis (rteague@redhat.com)
flake8 mccabe dependency fix (rteague@redhat.com)
Temporary work-around for flake8 vs maccabe version conflict
More toxification (jdetiber@redhat.com)

Wait for nodes to be ready before proceeding with upgrade.
Add 10 second wait after disabling firewalld (sdodson@redhat.com)

g_master_mktemp in openshift-master conflicts with
Use openshift.common.hostname when verifying API port available.
Add a fact to select --evacuate or --drain based on your OCP version

Automatic commit of package [openshift-ansible] release [3.4.54-1].
Ensure serial certificate generation for node and master certificates.
Create individual serving cert and loopback kubeconfig for additional

Update example sync script (sdodson@redhat.com)
Update v1.4 content, add api-gateway (sdodson@redhat.com)
Support openshift_node_port_range for configuring service NodePorts
Move role dependencies out of playbooks for openshift_master, openshift_node
should be libvirt. :-) (zhang.wanmin@zte.com.cn)
Fix repo defaults (sdodson@redhat.com)
Use registry.access.redhat.com/rhel7/etcd instead of etcd3

Revert "Deprecate node 'evacuation' with 'drain'" (sdodson@redhat.com)
Re-add when condition which was removed mistakenly in #3036
Fix jsonpath expected output when checking registry volume secrets
logging-deployer pull fixes from origin-aggregated-logging/#317
Don't upgrade etcd on atomic host, ever. (sdodson@redhat.com)
Fix jsonpath expected output when checking registry volume secrets
Enable repos defined in openshift_additional_repos by default
Hush the sudo privs check in oo-installer (tbielawa@redhat.com)
Add future versions to openshift_facts (ccoleman@redhat.com)
Cast openshift_enable_origin_repo to bool. (abutcher@redhat.com)
Update CFME template to point to GA build (simaishi@redhat.com)
Add templates for CFME Beta pod images (simaishi@redhat.com)
Add osnl_volume_reclaim_policy variable to nfs_lvm role
remove duplicate filter name and oo_pdb (jdetiber@redhat.com)
remove old Ops tooling (jdetiber@redhat.com)
enable pip cache for travis (jdetiber@redhat.com)
python3 support, add tox for better local testing against multiple python
modify_yaml: handle None value during update. (abutcher@redhat.com)
Update the openshift-certificate-expiry README to reflect latest changes
Deprecate node 'evacuation' with 'drain' (tbielawa@redhat.com)
Add master config hook for 3.4 upgrade and fix facts ordering for config hook
The next registry.access.redhat.com/rhel7/etcd image will be 3.0.15
[uninstall] Remove excluder packages (sdodson@redhat.com)
Check embedded etcd certs now, too (tbielawa@redhat.com)
Include 'total' and 'ok' in check results (tbielawa@redhat.com)
Fix access_modes initialization (luis.fernandezalvarez@epfl.ch)
Updated OpenShift Master iptables rules (rteague@redhat.com)
YAML Linting (rteague@redhat.com)
Make both backup and upgrade optional (sdodson@redhat.com)
[upgrades] Upgrade etcd by default (sdodson@redhat.com)
upgrades - Fix logic error about when to backup etcd (sdodson@redhat.com)
Limit node certificate SAN to node hostnames/ips. (abutcher@redhat.com)
Make 'cover-erase' a config file setting. Move VENT target to pre-req for all
Fixes to 'make ci' (tbielawa@redhat.com)
Resolved lint issues (rteague@redhat.com)
Minimum Ansible version check (rteague@redhat.com)
Removed verify_ansible_version playbook refs (rteague@redhat.com)
Fix coverage not appending new data (tbielawa@redhat.com)
Drop 3.2 upgrade playbooks. (dgoodwin@redhat.com)
Silence warnings when using rpm directly (dag@wieers.com)
Silence warnings when using rpm directly (dag@wieers.com)
Silence warnings when using rpm directly (dag@wieers.com)
Remove Hostname from 1.1 and 1.2 predicates (jdetiber@redhat.com)
Properly handle x.y.z formatted versions for openshift_release
etcd_upgrade: Simplify package installation (sdodson@redhat.com)
Speed up 'make ci' and trim the output (tbielawa@redhat.com)
add comments and remove debug code (jdetiber@redhat.com)
Pre-pull master/node/ovs images during upgrade. (dgoodwin@redhat.com)
Handle updating of scheduler config during upgrade (jdetiber@redhat.com)
Fix templating (jdetiber@redhat.com)
test updates (jdetiber@redhat.com)
Always install latest etcd for containerized hosts (sdodson@redhat.com)
etcd_upgrade : Use different variables for rpm vs container versions
Switch back to using etcd rather than etcd3 (sdodson@redhat.com)
node_dnsmasq - restart dnsmasq if it's not currently running
Conditionalize master config update for admission_plugin_config.
upgrade_control_plane.yml: systemd_units.yaml nees the master facts
openshift-master/restart : use openshift.common.hostname instead of
Update scheduler predicate/priorities vars (jdetiber@redhat.com)
fix tags (jdetiber@redhat.com)
openshift_node_dnsmasq - Remove strict-order option from dnsmasq
Fix metricsPublicURL only being set correctly on first master.
Explictly set etcd vars for byo scaleup (smunilla@redhat.com)
Cleanup ovs file and restart docker on every upgrade. (dgoodwin@redhat.com)
Sync latest image stream and templates for v1.3 and v1.4 (sdodson@redhat.com)
xpaas v1.3.5 (sdodson@redhat.com)
Ansible version check update (tbielawa@redhat.com)
allow 'latest' origin_image_tag (sjenning@redhat.com)
Remove duplicate when key (rteague@redhat.com)
refactor handling of scheduler defaults (jdetiber@redhat.com)
update tests and flake8/pylint fixes (jdetiber@redhat.com)
fix tagging (jdetiber@redhat.com)
do not report changed for group mapping (jdetiber@redhat.com)
fix selinux issues with etcd container (dusty@dustymabe.com)
etcd upgrade playbook is not currently applicable to embedded etcd installs
Fix invalid embedded etcd fact in etcd upgrade playbook.
Gracefully handle OpenSSL module absence (misc@redhat.com)
Refactored to use Ansible systemd module (rteague@redhat.com)
Updating docs for Ansible 2.2 requirements (rteague@redhat.com)
Fix the list done after cluster creation on libvirt and OpenStack
Set nameservers on DHCPv6 event (alexandre.lossent@cern.ch)
Systemd `systemctl show` workaround (rteague@redhat.com)
Verify the presence of dbus python binding (misc@redhat.com)
Update README.md (jf.cron0@gmail.com)
Reference master binaries when delegating from node hosts which may be
Merge kube_admission_plugin_config with admission_plugin_config
Added a BYO playbook for configuring NetworkManager on nodes
Make the role work on F25 Cloud (misc@redhat.com)
Make os_firewall_manage_iptables run on python3 (misc@redhat.com)
Modified the error message being checked for (vishal.patil@nuagenetworks.net)
Only run tuned-adm if tuned exists. (dusty@dustymabe.com)
Delegate openshift_manage_node tasks to master host. (abutcher@redhat.com)
Fix rare failure to deploy new registry/router after upgrade.
Refactor os_firewall role (rteague@redhat.com)
Allow ansible to continue when a node is unaccessible or fails.
Create the file in two passes, atomicly copy it over (sdodson@redhat.com)
Escape LOGNAME variable according to GCE rules (jacek.suchenia@ocado.com)
node_dnsmasq -- Set dnsmasq as our only nameserver (sdodson@redhat.com)
Refactor to use Ansible package module (rteague@redhat.com)
Allow users to disable the origin repo creation (sdodson@redhat.com)
Fix yum/subman version check on Atomic. (dgoodwin@redhat.com)
Check for bad versions of yum and subscription-manager. (dgoodwin@redhat.com)
Corrected syntax and typos (rteague@redhat.com)
Fix GCE cluster creation (lhuard@amadeus.com)
Optimize the cloud-specific list.yml playbooks (lhuard@amadeus.com)
Added ip forwarding for nuage (vishal.patil@nuagenetworks.net)
Add view permissions to hawkular sa (sdodson@redhat.com)
Use multiple '-v's when creating the metrics deployer command
Change to allow cni deployments without openshift SDN (yfauser@vmware.com)
[openstack] allows timeout option for heat create stack
Add hawkular admin cluster role to management admin (fsimonce@redhat.com)
Prevent useless master by reworking template for master service enf file
support 3rd party scheduler (jannleno1@gmail.com)
Add nuage rest server port to haproxy firewall rules. (abutcher@redhat.com)
storage/nfs_lvm: Also export as ReadWriteOnce (walters@verbum.org)

logging-deployer pull fixes from origin-aggregated-logging/#317

Don't upgrade etcd on atomic host, ever. (sdodson@redhat.com)
Update CFME template to point to GA build (simaishi@redhat.com)
Don't upgrade etcd on atomic host, ever. (sdodson@redhat.com)
Add templates for CFME Beta pod images (simaishi@redhat.com)

Revert "Pre-pull master/node/ovs images during upgrade."

modify_yaml: handle None value during update. (abutcher@redhat.com)

Add master config hook for 3.4 upgrade and fix facts ordering for config hook
The next registry.access.redhat.com/rhel7/etcd image will be 3.0.15

Make both backup and upgrade optional (sdodson@redhat.com)
[upgrades] Upgrade etcd by default (sdodson@redhat.com)

etcd_upgrade: Simplify package installation (sdodson@redhat.com)

Pre-pull master/node/ovs images during upgrade. (dgoodwin@redhat.com)
Properly handle x.y.z formatted versions for openshift_release
Remove Hostname from 1.1 and 1.2 predicates (jdetiber@redhat.com)
add comments and remove debug code (jdetiber@redhat.com)
Handle updating of scheduler config during upgrade (jdetiber@redhat.com)