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
* Defer loading result_stdout_text until specifically needed
* Conditionally display it based on the size of the field
* Provide a helpful message otherwise
This commit makes it so that enterprise licenses are rejected in HA
environments if there is no active MongoDB server.
Additionally, it suppresses trying to connect to MongoDB in cases where
it is not present or meaningful.
The flake8 command was identifying several warnings and errors. This change
addresses the flake8 warnings and updates the setup.cfg with additional
exclusions. If accepted, jenkins will be updated to use the flake8 command,
rather than using the django_jenkins plugin. This will expedite jenkins
testing.
* openstack_inventory_support:
Use inventory_source source_vars to control what address we use for ansible_ssh_host
Some small changes to tower openstack work
1) Favor private addresses in the openstack inventory sync
2) Force refresh
3) use hostId and promote it to the top level vars object so we can
track individual host changes
* added mongo connection logic
* added mongo dbtransform logic to allow keys with . and $
* altered tower fact scanner CacheModule to emit a message for each fact module facts (including ansible facts). Previously, seperate facts module facts were getting concatenated to each subsequent emi
* tower fact scanner CacheModule timeout set as to not hang for forever
* broke apart commands.py test
* added unit test for run_fact_cache_receiver, facts, and dbtransform
* Add logic to ansible callback plugin to prevent it from waiting
forever to submit events to Tower
* Lower process recycle threshold for tower callback receiver
* Make recycle threshold configurable
* Properly exit the main callback receiver management process if
the event receiver process is dead so we don't leave dead worker
processes
* Set a configurable maximum number of messages that can be waiting
in a worker process queue before it is skipped instead of filling
up memory on a dead worker process
* Skip over a dead worker process if it's queue is full
* Force restart callback receiver if all queues are dead
* Roll back transaction.atomic with the thought that it is causing
deadlocks in the worker process. Use the old commit_on_success
mechanism with retry logic
* Seperate queue nonblocking expected exception from any other type
of exception that could be encountered on the queue fetch operation
* origin/master:
Add tests for inventory_filters and group_by, fix to convert None to empty string.
Job submission
Fix up unified job creation parameter passing
Roll this back, it doesn't do what I expect
Make sure we propogate source_script into the related fields it needs to be in
Update job template launch docs with credential_needed_to_start documentation
Custom Inventory Scripts
forms/Sources.js
Add instance filters and group by options for EC2 inventory sources. Implements https://trello.com/c/QOVhP0mH
Make sure when we call create_unified_job we can pass important fields as _id or without
Make sure we pass extra parameters down to the unified job create method so we can pick up credential if given
Add system job template launch documentation
Track source_script in summary and related fields for relevant inventory sources
Host events
Simplify include_vars based on mpd feedback
Update the apt-cache when installing python-pip
Only the primary should syncdb and migrate
Update package changelogs
Update release history and process docs
Update ansible-tower service config in playbook
