* awx.main.models Fact added
* view host fact and timeline updated to use new Postgres Fact model instead of Mongo
* Removed license set start Mongo logic
* added View tests
* added Model tests
* Removed mongo fact unit tests
* point at modified jsonbfield that supports sqlite storage driver
* postgresify fact cache receiver
* test OPTIONS endpoint
* Note: single fact view not implemented yet.
* Job object can now control the output and generate K:V output for
notification types that can support it
* Notifications store the body as json/dict now to encode more
information
* Notification Type can further compose the message based on what is
sensible for the notification type
* This will also allow customizing the message template in the future
* All notification types use sane defaults for the level of detail now
* Remove fields from browsable API raw data that are set implicitly based on URL / parent object.
* Fix issue where a group/host could be assigned to a different inventory.
* Update validation to load values from existing instance if not present in new data; allows PATCH requests to succeed.
* Remove job_args, job_cwd, job_env, result_stdout and result_traceback fields from job listings.
* Implement concrete Notification model for notification runs
* Implement NotificationTemplate and Notification serializers and views
* Implement ancillary views
* Implement NotificationTemplate trigger m2m fields on all job templates
via a fields mixin
* Link NotificationTemplates with an org
* Link notifications with the activity stream
* Implement Notification celery tasks
* Extend Backend field parameters to identify sender and receiver as
parameters needed by the message and not the backend itself
* Updates to backends to better fit the django email backend model as it
relates to Messages
* Implement success job chain task + notifications
* Implement notifications in error job chain task
* Fixing some bugs in the manifest definition
* Database model and schema migration for tower settings
* Initial View and Serializer implementation using a strategy of merging
model instances and named tuples
* Move auth backends into sso app.
* Add support for mapping social auth users into organizations and teams.
* Return social auth backends in a consistent order in the API.
* Remove custom SAML attribute mapping and use options provided by PSA.
* Add pipeline function to raise an exception if no user has been found or created; added comments on how to disable new user creation.
* Add comments for defining a custom social auth pipeline function.
* content_format=ansi&content_encoding=base64 support added to
?format=json on stdout endpoint.
Currently the json `content` value is made into html.
This change will allow support for consuming STDOUT for display on the
terminal.
* upstream/release_2.3: (91 commits)
Include python-{paramiko,ecdsa} dependencies
Remove extra epel testing stanzas
Unit test for ec2 credentialless inventory
Fix issue with ec2 iam sync with no credential.
Use the htpasswd command instead the ansible module
Pip is no longer needed
check local user root or not in ./configure
Remove unneeded when check for super user addition
Improve distro detection in setup.sh
Fix superuser check on upgrade
Minor improvements to setup.sh
Remove ansible prerequisite check from configure
Attempt to install ansible within setup.sh
Allow munin processes to access postgres
Move up base package dependency install
fixes jenkins failures
Proper flake8 fix
fixes executing processes with correct PYTHONPATH will pickup .pth files
Show the repo for bundled package file dump
Proper flake8 fix
...
* inherit extra_vars from job_template
* overwrite with extra vars survey default vars
* overwrite with job extra_vars
* bypass extra_vars inherit logic for job relaunch
* with tests
The act of committing a PR that including part of these changes and then
reverting it meant that some of the changes from the original branch
didn't land in this new PR. This commit adds the missing bits.
Lesson learned: Create a new branch when submitting a PR on a
previously reverted PR
* Cleanup old code related to stdout download using temporary files
* Remove copy-into-database code
* Modify stdout download code to reference stdout file
The act of committing a PR that including part of these changes and then
reverting it meant that some of the changes from the original branch
didn't land in this new PR. This commit adds the missing bits.
Lesson learned: Create a new branch when submitting a PR on a
previously reverted PR
* Cleanup old code related to stdout download using temporary files
* Remove copy-into-database code
* Modify stdout download code to reference stdout file
