* Adding integration test for 127.0.0.1/32 and ::1/128.
* Making sure file is not corrupted when render fails
* Fixes#56430
* Adding changelog for MR 57147/Issue 56430
Set ansible_buffer_read_timeout to be larger then
ansible_command_timeout, as both appliance / controller are on the same
physical network and we are not triggering ansible_command_timeout.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
* ios_ntp: Switch interface to GigabitEthernet0/1 for tests
Lets use an interface we know that exists for testing, in some cases
Loopback0 maybe not be configured.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
* ios_ntp: properly register result variable for testing
This is currently broken, and tests don't work. Fix this by properly
registering the result variable.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
ios_user requires escalated privledges, rather then doing this in the
inventory, we can just add it into the playbook tasks.
Also add missing provider settings for authorize.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
This was required to be installed by a human, out side of our testing.
Add it so ansible-test can now manage.
Signed-off-by: Paul Belanger <pabelanger@redhat.com>
* Updated testcase
* Added check mode support
* Added check for mutual exclusive for Name and UUID
Fixes: #57580
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
(cherry picked from commit 00604d3f2c)
To get all instances gcp_compute made a call to the Google API for each
zone separately. Because of this if all zones needed to be queried
fetching hosts lasted 30+ seconds. Now the module will use a single
query that will return all the instances, so the execution should last
just a few seconds.
This commit also suppresses a warning from the google-auth library about
using user credentials because if an Ansible user wants to use user
credentials, there is no need to warn him about it.
(cherry picked from commit f6a0f9874d)
* preserve same order as inventory manager when using host lookup (#55331)
* preserve same order as inventory manager when using inventory_hostnames lookup
add a test
* move generic code
(cherry picked from commit afb5e02c19)
* changelog
* Found bug, fixed by moving the serialization of objects out of try while creating objects
* Added changelog to document fix
(cherry picked from commit d07d394779)
* update ce_ntp.py and remove the root tag name to find all nodes (#56976)
* add a maintainer
* update ce_ntp.py and remove the root tag name to find all node
* update ce_ntp.py and remove the root tag name to find all node
(cherry picked from commit 0137c4343b)
* add changelogs fragments
* add changelogs fragments
The extant documentation says that the fingerprint return value is a
single string, but it is currently being returned as a split list.
Convert the returned value to a string as documented, and add some
basic test-case coverage for the return values.
(cherry picked from commit 6f06fc9945)
* win_chocolatey - honour version when bootstrapping chocolatey and fix package listing (#57208)
* win_chocolatey - honour version when bootstrapping chocolatey
* skip upgrade all step
* Fix install latest step
* Remove test changes now that Chocolatey is released
* tweak the package version detection
(cherry picked from commit b5b75daaa5)
* removed disabled alias
* nxos_vlan: fix broken purge behavior (issue #57101) (#57229)
* nxos_vlan: fix broken purge behavior (issue #57101)
Symptoms/Analysis:
- `nxos_vlan` `purge: true` would fail when `purge` was trying to delete all unspecified vlans, including vlan 1.
- `nxos` devices do not allow removing vlan 1 and raise a cli exception error
- Previous fix#55144 caused a side effect when `purge` was used: vlan changes specified by `aggregate` were ignored; e.g.
- vlan 4 is not present; playbook specifies `aggregate: { vlan: 4 }, purge: true`
- results in proper purging but vlan 4 is not created
Solutions:
- ignore vlan 1 when purging
- remove the `not purge` check from state present logic
Added additional unit tests and integration tests.
Tested against all regression platforms.
* PEP fixes
* Add agg_show_vlan_brief.txt fixture
* Add warning for removing vlan 1
* change method name check
(cherry picked from commit 6bb13bbb84)
* changelog
Signed-off-by: Trishna Guha <trishnaguha17@gmail.com>
* Fix order for warning on templated conditionals
Fix bare variable warnings when the variable is a boolean
* changelog
* Add tests for cases that should and should not give warnings
If the behavior may change when the default behavior for CONDITIONAL_BARE_VARS becomes False there should be a warning. Boolean type conditionals will not change in behavior so don't warn.
* oops, forgot to add files
* typo
(cherry picked from commit 21cd24a0dd)
According to the OpenStack Networking API
the attribute binding:vnic_type of a port is optional.
This change enables the os_port module to handle
binding:vnic_type as optional.
(cherry picked from commit bc50a52ee2)
Stripping the output from `podman image ls -q` prevents passing
a blank image ID to `podman image inspect`, which would fail.
(cherry picked from commit a05be890ad)
Co-authored-by: Matt Clay <matt@mystile.com>
Changes example to `when: inventory_hostname == ansible_play_hosts_all[0]` to really run run_once regardless of serial.
(cherry picked from commit dea9644d21)
The erlang-solutions repository is broken for Ubuntu18 (They did not
sign their repository metadata). For now, disable the rabbitmq tests
which depend upon that. I'll open a PR with a revert of this commit.
We can watch it to see when it passes in Ci to know that the
erlang-soutions repository has been fixed
(cherry picked from commit 5f47ab9)
Co-authored-by: Toshio Kuratomi <a.badger@gmail.com>
