| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Ansible 2.4
|
|
|
|
|
|
|
| |
openshift.master.cluster_method.
With pacemaker removed there is no longer a need for
openshift.master.cluster_method. We only have one option.
|
| |
|
| |
|
|\
| |
| |
| |
| | |
Automatic merge from submit-queue.
preserve the oo-install ansible_inventory_path value
|
| | |
|
| |
| |
| |
| | |
Previous installer release v3.5 used to display the 3 variants.
I keep OSE as the default, but proposes to add back the "origin v3.6" variant.
|
| |
| |
| |
| |
| |
| | |
Fixes bugzilla : https://bugzilla.redhat.com/show_bug.cgi?id=1460353
Signed-off-by: jkaurredhat <jkaur@redhat.com>
|
|\ \
| |/
|/| |
Fix BZ#1422541 BZ#1440586 BZ#1440612 BZ#1440617 on master branch
|
| | |
|
| |
| |
| |
| | |
Fixes bug 1447204
|
| | |
|
|/ |
|
|
|
|
|
|
|
|
|
|
| |
Fixes issues dumping various objects to YAML. This change updates the
quick installer fact caching callback to use the improved
AnsibleDumper when writing out the facts to storage.
* https://bugzilla.redhat.com/show_bug.cgi?id=1421053
* https://bugzilla.redhat.com/show_bug.cgi?id=1421061
* https://bugzilla.redhat.com/show_bug.cgi?id=1421059
|
| |
|
|
|
|
|
| |
Found using `vulture`, confirmed that code that was once used became
unused in later commits after its introduction.
|
|
|
|
| |
3.5 added and 3.4 moved to LEGACY
|
| |
|
|
|
|
|
|
| |
* 'atomic-openshift-installer' will exit now if -u is given with the
'scaleup' subcommand
* https://bugzilla.redhat.com/show_bug.cgi?id=1390135
|
|
|
|
|
|
|
| |
* Don't print words when checking for sudo access
* Just 'echo' an empty string with no new-line
Closes #3003
|
|
|
|
| |
versions
|
| |
|
|\
| |
| | |
Enable quiet output for all a-o-i commands
|
| |
| |
| |
| |
| |
| |
| | |
* Update: uninstall, upgrade, scaleup wf's
* Verbosity can be toggled with -v flag, as before
https://bugzilla.redhat.com/show_bug.cgi?id=1384294
|
|/
|
|
|
|
| |
Update the default for the override prompt and set both
openshift_master_cluster_hostname and
openshift_master_cluster_public_hostname
|
| |
|
|
|
|
|
|
| |
Just a quick addition for the upgrade playbooks for 3.4
Fixes Bug 1384291
|
|\
| |
| | |
Update master_lb vs cluster_hostname workflow
|
| |
| |
| |
| |
| |
| | |
Previously, the messaging around setting cluster_hostname
for single master setups was confusing. This should help with
that as well as removing unecessary values from the cfg file.
|
| |
| |
| |
| |
| |
| | |
* Load balancers are now excluded from the mixed environment check
* https://bugzilla.redhat.com/show_bug.cgi?id=1390064
|
| |
| |
| |
| |
| |
| |
| |
| | |
We have changed how we assign etcd members in the quick
installer. This updates our messaging to match those changes.
Fixes Bug 1383961
Fixes Bug 1389649
|
|/
|
|
|
|
| |
Update the quick installer to correctly set the new version.
Fixes Bug 1388762
|
|\
| |
| | |
a-o-i: Separate install and scaleup workflows
|
| |
| |
| |
| |
| |
| |
| | |
In order to simplify the quick installer logic, this moves
the scaleup workflow into its own subcommand.
Fixes: Bug 1339621
|
|/ |
|
|
|
|
|
|
|
| |
Update to use the correct playbook for 3.2 to 3.3 upgrades
and some error handling for unsupported upgrades.
Fixes Bug 1381411
|
| |
|
|
|
|
|
|
|
| |
If uninstalled hosts were in detected in an environment,
the installer could get into a hung state that a user
could not easily recover from. This lets --force do what is
more expected.
|
|
|
|
|
| |
Handle openshift_node_labels separately because they
need to be doublequoted.
|
|
|
|
|
| |
Allow the user to set arbitrary variables directly under the host item instead of
in under an 'other_variables' item.
|
|
|
|
| |
Also, remove the final 'press a key to continue' prompt.
|
| |
|
| |
|
|\
| |
| | |
a-o-i: Do not display version number in quick installer
|
| |
| |
| |
| |
| |
| | |
As we now require the version of the installer to match what will be installed in
the target environment and to avoid any potential confusion, remove the version
number from the variant choice question.
|
|/
|
|
|
|
| |
Make sure we don't set an unschedulable node as infra as that can cause problems.
Fixes: Bug 1375946
|
| |
|
|\
| |
| | |
Fix PyLint errors discovered when upgrading to newer version
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Fixes PyLint to run in the virtualenv used for all tests
* Replaced 'LooseVersion' with 'parse_version' from setuptools
- This is a work around for the issue in
https://github.com/PyCQA/pylint/issues/73 in which pylint can not
import disutils.version correctly in a virtualenv.
* Removed the unused function 'delete_hosts' which was causing a
pylint error as well
* Removed a deprecated pylint pragma option, 'bad-builtin'
* Fixed some import ordering issues it was picky about
* Added another disable for a case where the PyLint suggestion would
have us altering the container we would be iterating over
* Add code-coverage reports to the unittests with the MINIMUM coverage
percentage for success set to 70%
- Current test coverage is at 76%
|