* templar: ensure that exceptions are handled
* Fix AttributeError: object has no attribute 'message'
'message' attribute is deprecated since Python 2.6 and not available
with Python 3.
Simple reproducer:
- hosts: localhost
vars:
not_json: "{{ 'test str' | from_json }}"
tasks:
- command: "echo {{ not_json }}"
* ignore empty lines in rabbitmqctl output
this fixes a bug with rabbitmq 3.7.5
rabbitmqctl can return empty lines, breaking the rabbitmq_parameter module
especially in a new vhost, the command rabbitmqctl list_parameters -q -p <vhost> will return an empty line
* Strip empty line in rabbitmqctl output
* Fix iosxr netconf plugin response namespace
* iosxr netconf plugin removes namespace by default
for all the responses as parsing of xml is easier
without namepsace in iosxr module. However to validate
the response received from device against yang model requires
namespace to be present in resposne.
* Add a parameter in iosxr netconf plugin to control if namespace
should be removed from response or not.
* Fix CI issues
* Fix review comment
* nxos_interface vlan and port-channel idempotence fix for mtu
* Fix MTU reconfiguration at each execution
* nxos_interface port-channel idempotence fix for mode (#44248)
* Fix trunk mode idempotence for port-channel
* Gather ethernet and port-channel code for mode management
* nxos_linkagg port-channel idempotence fix for channel-group's mode
The regex to retrieve channel-group's mode is not enough accurate.
Therefore, the swhitchport mode was matched instead of the
channel-group's mode.
* This fix add accuracy to match the right configuration command
* Add support for switchport mode dot1q-tunnel in nxos_interface
* Fix layer reconfiguration at each execution
* require git when trying to use it in ansible-galaxy cli
Previously we weren't setting `required=True` when calling
`get_bin_path` and the path would return `None`, this would cause
a traceback when attempting to ' '.join() to create a string
representation of the failed command for error output
Fixes#49200
Signed-off-by: Adam Miller <admiller@redhat.com>
* add changelog
Signed-off-by: Adam Miller <admiller@redhat.com>
* Changes to redfish-based parameters
Corrected (potentially) small scope variable to 'id'
Review to see if this is problematic
* Updated doc information with "version_added"
* Corrected 'username' in redfish_config file
- The role name and instance profile name _can_ be different
- Change the delimiter to `:` for keys that are discovered through the JSON parsing (which is not a valid delimiter for AWS IAM role names), this delimiter is still converted to underscore
- Now checks for the existence of that delimiter to remove the cases where the JSON keys are appended to the role name to find the role name
Different connection plugins return different data when throwing exceptions. The Paramiko connection plugin does not return a text sting, which caused an exception.
The ssh connection plugin returns multi-line errors, which makes the debug logs harder to read. Only return the last line in that case in order to make the logs more readable.
When experiencing a connection failure, reset the connection.
Add reset() to paramiko_ssh
Indicate thet conection state is False when running close(). This is needed by the ensure_connected() decorator to work properly.
Co-authored-by: Matt Martz <matt@sivel.net>
* test for openstack inventory constructed functionality
this adds unit tests for the compose, groups, and keyed_var features
of the openstack inventory plugin
* fix constructed functionality in openstack inventory plugin
The compose, groups, and keyed_groups functionality of the openstack
inventory plugin was broken:
- the plugin was not passing the correct variables to the
Constructable methods for compose and groups
- the plugin was simply never calling the appropriate method for
implementing keyed_groups
This commit fixes both issues.
Network platforms that don't have cliconf plugin will fail when
sending rpc calls for the reset_history and disable_response_logging
functions because those are defined in cliconf exclusively at this
time.
This patch adds checks for those attributes before making the call
* Don't check options for idempotency which are not supported.
This check should be superfluous if every option would adhere to
the convention that options not specified should have value None.
Unfortunately, some options (such as init) which correspond to
container properties have an explicit default set.
Path and Port are mutually exclusive parameters but not documented.
This fix documents this requirement.
Fixes: #15732
Signed-off-by: Abhijeet Kasurde <akasurde@redhat.com>
* Added win_partition module
* Fixes sanity tests
* Requested changes minus changes in partition_size
* Removed trailing whitespace and fixed docstring
* Changes to parititon_size to allow variable units
* Stricter regex for partition_size
* Updated help to list all features
just to avoid searching powershell syntax Get-WindowsFeature
+label: docsite_pr
* added Linebreak
added Linebreak
* removed trailing whitespace
removed trailing whitespace
* slight modification of text
* Support UpdateService forceNewDeployment in ecs_service module
* Force update to be called if force_new_deployment set
* Fixes for review
* Add force_new_deployment option to ecs_service.py
cherrypicks changes from via/ansible
Adds tests for pull request #42518
fixes backwards compatability with boto<1.8.4
* change version_added to 2.8 for force_new_deployment
* remove extra lines from test
* remove more unnecessary whitespace
* Update docs with rhsm note
redhat_subscription is the module for handling registration of modern Red Hat systems. This commit adds a note indicating that unless people are using RHEL 5 etc, they probably want that module instead.
+label: docsite_pr
* Make redhat_subscription a link to the module docs
* Added feature health_check_grace_period_seconds to ecs_service, this time with a botocore version check and some initial testing
* Only set health_check_grace_period_seconds when loadbalancers are defined
* Removed leftover commas and fix in test
* Removed blank line
* Minor improvements for ecs_service module
* Removed default (30) for health_check_grace_period_seconds param
* Changed botocore version allowed to 1.8.20 for health check param.
* Fix empty healthcheck failure
* nxos_facts: Remove dead code
The commit e51964e made this redundant as the structured case is handled
elsewhere.
* nxos_facts: Do not gather neighbors redundantly
LLDP reports the neighbor using the abbreviated interface name, whereas
CDP reports the neighbor using the full interface name. Normalize the
local interface name in the LLDP case, so there is no redundant
information. Due to the order of the gathering, CDP neighbors are saved
in case both LLDP and CDP data is available on a certain interface.
Releases can be listed without logging in for public projects, so allow
`github_release` to be called without `token` or `password`.
Signed-off-by: Benoît Knecht <benoit.knecht@fsfe.org>
* Fixes#38174: Add remove-brick operation for gluster_volume
Existing gluster_volume module does not support removal of bricks
from a gluster volume. Also, operation remove-brick needs to be
followed by a commit operation.
Signed-off-by: Devyani Kota <dkota@redhat.com>
* Updates #38174: Add documentation for remove-brick operation
Signed-off-by: Devyani Kota <dkota@redhat.com>
* Updates #38174: Add support to reduce cluster configuration.
An exceptional scenario exists where the user might want to
reduce the gluster cluster configuration for replicated
volumes from replica 3 to replica 2 that need to be handled
in a different manner than the generic gluster commands,
where the user is expected to mention the replica count
of the new configuration in order to remove the bricks.
Signed-off-by: Devyani Kota <dkota@redhat.com>
* Updates #38174: Add checks for self-heal status prior remove operation.
While reducing the cluster configuration from replica 3
to 2, it needs to be checked for status of self-heal
prior remove-brick operation is executed.
Signed-off-by: Devyani Kota <dkota@redhat.com>
* Add append_hash functionality to k8s module
append_hash adds a hash based on the contents of a ConfigMap
or Secret to the name - this enables immutable ConfigMaps and
Secrets.
* Provide k8s_config_resource_name plugin
The k8s_config_resource_name filter plugin provides a means of determining
the name of ConfigMaps and Secrets created with append_hash
* Add changelog fragment
* fix failing tests
* Update openshift version needed for append_hash
* autostart cannot be used with command=define
autostart cannot be used with command=define. If it's used with state, it'll be silently ignored.
https://github.com/ansible/ansible/issues/41592
* guest name should not be used with xml
When both ```name``` and ```xml``` are specified, there might be a mismatch of domain names in
```name``` and inside ```xml```.
* Improved error handling and param deps for autostart-only tasks.
Standalone autostart tasks depend on ```name``` and need the domain to be present.
* Added handling of errors thrown by libvirt
* Updates to documentstion and examples
* Removed required flag from name. Added description on some option combinations.
Added a few examples.
* lint issues
* docs: note that name is optional
* Removing required from documentation of name parameter
* extra text to make clearer when name is required
* When defining a domain with xml, its name is taken directly from the xml definition.
This reverts commit 4ac14a622b.
* win_mapped_drive - refactor module and docs
* Updated code to work with become and split tokens
* use win_credential_manager instead of cmdkey
* updated credential manager module name
* harden the system token impersonation process
