* Restructure roadmap so that we can link to sub-parts
So we want to point to specific subsections in the roadmap from the Working Group pages.
This requires the use of subtitles rather than a long list of items and sub-items.
* RST syntax is hard...
* Fixes as requested
* Another fix
* Converted all ROADMAPs to new format
- Figuring out privilege escalation (runas w/ username/password)
- Implement kerberos encryption over http
- pywinrm conversion to requests (Some mess here on pywinrm/requests. will need docs etc.)
Windows
- NTLM support
-------
- General
- Figuring out privilege escalation (runas w/ username/password)
- Implement kerberos encryption over http
- pywinrm conversion to requests (Some mess here on pywinrm/requests. will need docs etc.)
- NTLM support
- Modules
- Modules
- Windows
- Finish cleaning up tests and support for post-beta release
- Finish cleaning up tests and support for post-beta release
- Strict mode cleanup (one module in core)
- Strict mode cleanup (one module in core)
- Domain user/group management
- Domain user/group management
- Finish win_host and win_rm in the domain/workgroup modules.
- Finish win_host and win_rm in the domain/workgroup modules.
- Close 2 existing PRs (These were deemed insufficient)
- Replicate python module API in PS/C# (deprecate hodgepodge of stuff from module_utils/powershell.ps1)
- Close 2 existing PRs (These were deemed insufficient)
- Network
- Replicate python module API in PS/C# (deprecate hodgepodge of stuff from module_utils/powershell.ps1)
- Cisco modules (ios, iosxr, nxos, iosxe)
- Arista modules (eos)
Network
- Juniper modules (junos)
-------
- OpenSwitch
- Cisco modules (ios, iosxr, nxos, iosxe)
- Cumulus
- Arista modules (eos)
- Dell (os10) - At risk
- Juniper modules (junos)
- Netconf shared module
- OpenSwitch
- Hooks for supporting Tower credentials
- Cumulus
- VMware (This one is a little at risk due to staffing. We're investigating some community maintainers and shifting some people at Ansible around, but it is a VERY high priority).
- Dell (os10) - At risk
- vsphere\_guest brought to parity with other vmware modules (vs Viasat and 'whereismyjetpack' provided modules)
- Netconf shared module
- VMware modules moved to official pyvmomi bindings
- Hooks for supporting Tower credentials
- VMware inventory script updates for pyvmomi, adding tagging support
- Azure (Notes: This is on hold until Microsoft swaps out the code generator on the Azure Python SDK, which may introduce breaking changes. We have basic modules working against all of these resources at this time. Could ship it against current SDK, but may break. Or should the version be pinned?)
VMware
- Minimal Azure coverage using new ARM api
------
- Resource Group
This one is a little at risk due to staffing. We're investigating some community maintainers and shifting some people at Ansible around, but it is a VERY high priority.
- Virtual Network
- Subnet
- vsphere\_guest brought to parity with other vmware modules (vs Viasat and 'whereismyjetpack' provided modules)
- Public IP
- VMware modules moved to official pyvmomi bindings
- Network Interface
- VMware inventory script updates for pyvmomi, adding tagging support
- Storage Account
- Security Group
Azure
- Virtual Machine
-----
- Update of inventory script to use new API, adding tagging support
This is on hold until Microsoft swaps out the code generator on the Azure Python SDK, which may introduce breaking changes. We have basic modules working against all of these resources at this time. Could ship it against current SDK, but may break. Or should the version be pinned?)
- Docker:
- Minimal Azure coverage using new ARM api
- Start Docker module refactor
- Resource Group
- Update to match current docker CLI capabilities
- Virtual Network
- Docker exec support
- Subnet
- Upgrade other cloud modules or work with community maintainers to upgrade. (In order)
- Public IP
- AWS (Community maintainers)
- Network Interface
- Openstack (Community maintainers)
- Storage Account
- Google (Google/Community)
- Security Group
- Digital Ocean (Community)
- Virtual Machine
- Ansiballz (renamed from Ziploader):
- Update of inventory script to use new API, adding tagging support
- Write code to create the zipfile that gets passed across the wire to be run on the remote python
- Port most of the functionality in module\_utils to be usage in ansiballz instead
- Port a few essential modules to use ansiballz instead of module-replacer as proof of concept
Docker
- New modules will be able to use ansiballz. Old modules will need to be ported in future releases (Some modules will not need porting but others will)
------
- Better testing of modules, caching of modules clientside(Have not yet arrived at an architecture for this that we like), better code sharing between ansible/ansible and modules
- Start Docker module refactor
- ansiballz is a helpful building block for: python3 porting(high priority), better code sharing between modules(medium priority)
- Update to match current docker CLI capabilities
- ansiballz is a good idea before: enabling users to have custom module_utils directories
- Docker exec support
- Expand module diff support (already in progress in devel)
- Framework done. Need to add to modules, test etc.
Cloud
- Coordinate with community to update their modules
-----
- Things being kicking down the road that we said we’d do
Upgrade other cloud modules or work with community maintainers to upgrade. (In order)
- NOT remerging core with ansible/ansible this release cycle
- Community stuff
- AWS (Community maintainers)
- Define the process/ETA for reviewing PR’s from community
- Openstack (Community maintainers)
- Publish better docs and how-tos for submitting code/features/fixes
- Google (Google/Community)
- Digital Ocean (Community)
Ansiballz
---------
Renamed from Ziploader
- Write code to create the zipfile that gets passed across the wire to be run on the remote python
- Port most of the functionality in module\_utils to be usage in ansiballz instead
- Port a few essential modules to use ansiballz instead of module-replacer as proof of concept
- New modules will be able to use ansiballz. Old modules will need to be ported in future releases (Some modules will not need porting but others will)
- Better testing of modules, caching of modules clientside(Have not yet arrived at an architecture for this that we like), better code sharing between ansible/ansible and modules
- ansiballz is a helpful building block for: python3 porting(high priority), better code sharing between modules(medium priority)
- ansiballz is a good idea before: enabling users to have custom module_utils directories
Diff-support
------------
Expand module diff support (already in progress in devel)
- Framework done. Need to add to modules, test etc.
- Coordinate with community to update their modules
Other
-----
Things being kicking down the road that we said we’d do
- NOT remerging core with ansible/ansible this release cycle
Community
---------
- Define the process/ETA for reviewing PR’s from community
- Publish better docs and how-tos for submitting code/features/fixes
- Openshift: oso_deployment, oso_route, oso_service, oso_login (...and possibly others. These are modules being developed to support `ansible-container <https://github.com/ansible/ansible-container>`_.): Deferred for later release
- Kubernetes: kube_deployment, kube_service, kube_login (...and possibly others. These too are modules being developed to support `ansible-container <https://github.com/ansible/ansible-container>`_): Deferred for later release
- **Extras split from Core** (Team, Community, lead by Jason M and Jimi-c) (Targeting 2.2, could move into 2.3).
- Docker_network: **done**
Targeted towards the 2.2 release or shortly after, we are planning on splitting Extras out of the “Ansible Core” project. That means that modules that are shipped with Ansible by default are **only** the modules in ansibl-modules-core. Ansible extras will become a separate project, managed by the community standard. Over the next few months we’re going to have a lot of work to do on getting all of the modules in the right places for this to work.
- Docker_volume: Not in this release
- Docker_file: Not in this release.
- Openshift: oso_deployment, oso_route, oso_service, oso_login (...and possibly others. These are modules being developed to support `ansible-container <https://github.com/ansible/ansible-container>`_.): Deferred for later release
- Kubernetes: kube_deployment, kube_service, kube_login (...and possibly others. These too are modules being developed to support `ansible-container <https://github.com/ansible/ansible-container>`_): Deferred for later release
- Create proposal (Jason or Jimi)
Extras split from Core
- Review modules for correct location (extras v core)
----------------------
- Extras is a completely different package (does not install with ansible)
Lead by Jason M and Jimi-c (Targeting 2.2, could move into 2.3).
- Library dependencies
- Decide and implement release schedules between Ansible Core and Extras to ensure compatibility and longevity for modules and versions of Ansible.
- **Tweaks/Fixes**
Targeted towards the 2.2 release or shortly after, we are planning on splitting Extras out of the “Ansible Core” project. That means that modules that are shipped with Ansible by default are **only** the modules in ansibl-modules-core. Ansible extras will become a separate project, managed by the community standard. Over the next few months we’re going to have a lot of work to do on getting all of the modules in the right places for this to work.
- Connection handling stuff. (Toshio K. and Brian C.): This is a stretch goal for 2.2.**This work got pushed out**
- Create proposal (Jason or Jimi)
- Review modules for correct location (extras v core)
- Extras is a completely different package (does not install with ansible)
- Library dependencies
- Decide and implement release schedules between Ansible Core and Extras to ensure compatibility and longevity for modules and versions of Ansible.
- Change connection polling to avoid resource limitations, see `<https://github.com/ansible/ansible/issues/14143>`_
- vsphere: feature parity with whereismyjetpack and viasat modules
Windows
-------
Lead by Matt D
- Feature parity
- PS module API (mirror Python module API where appropriate). Note: We don’t necessarily like the current python module API (AnsibleModule is a huge class with many unrelated utility functions. Maybe we should redesign both at the same time?) (bumped to 2.3+ due to "moving target" uncertainty)
- PS module API (mirror Python module API where appropriate). Note: We don’t necessarily like the current python module API (AnsibleModule is a huge class with many unrelated utility functions. Maybe we should redesign both at the same time?) (bumped to 2.3+ due to "moving target" uncertainty)
- Environment keyword support(done)
- Environment keyword support(done)
@ -69,131 +91,149 @@ Target: September 2016
- Async support(done)
- Async support(done)
- (stretch goal) Pipelining(bumped to 2.3+)
- (stretch goal) Pipelining(bumped to 2.3+)
- **Windows-specific enhancements** (Matt D)
- Windows-specific enhancements
- Multiple Kerberos credential support (done)
- Multiple Kerberos credential support (done)
- Server 2016 testing/fixes(done, awaiting next TP/RTM)
- Server 2016 testing/fixes(done, awaiting next TP/RTM)
- (stretch goal) Nano Server connection + module_utils working (bumped to 2.3)
- (stretch goal) Nano Server connection + module_utils working (bumped to 2.3)
- (stretch goal) Encrypted kerberos support in pywinrm(bumped to 2.3)
- (stretch goal) Encrypted kerberos support in pywinrm(bumped to 2.3)
- **Network** (Nate C/Peter S)
Network
-------
- **Done:** Unify NetworkModules (module_utils/network.py) as much as possible
Lead by Nate C, Peter S
- **Done:** Add support for config diff and replace on supported platforms (2 weeks)
- **Done:** Support for VyOS network operating system
- **Done:** Unify NetworkModules (module_utils/network.py) as much as possible
- **Done:** Add support for RestConf for IOS/XE
- **Done:** Add support for config diff and replace on supported platforms (2 weeks)
- **Done:** Support for Dell Networking OS10
- **Done:** Support for VyOS network operating system
- Allow for multiple vault secrets https://github.com/ansible/ansible/issues/13243
- Add vault/unvault filters https://github.com/ansible/ansible/issues/12087 (deferred to 2.3)
- Config option to turn ‘unvaulting’ failures into warnings https://github.com/ansible/ansible/issues/13244
- Add vault support to lookups (likely deferred to 2.3 or until lookup plugins are revamped)
- Allow for multiple vault secrets https://github.com/ansible/ansible/issues/13243
- **Python3** (Toshio)
- Config option to turn ‘unvaulting’ failures into warnings https://github.com/ansible/ansible/issues/13244
A note here from Jason M: Getting to complete, tested Python 3 is both
a critical task and one that has so much work and so many moving parts
Python3
that we don’t expect this to be complete by the 2.2 release. Toshio will
-------
lead this overall effort.
Lead by Toshio
- Motivation:
A note here from Jason M: Getting to complete, tested Python 3 is both
- Ubuntu LTS (16.04) already ships without python2. RHEL8 is coming which is also expected to be python3 based. These considerations make this high priority.
a critical task and one that has so much work and so many moving parts
- Ansible users are getting restless: https://groups.google.com/forum/#!topic/ansible-project/DUKzTho3OCI
that we don’t expect this to be complete by the 2.2 release. Toshio will
- This is probably going to take multiple releases to complete; need to get started now
lead this overall effort.
- Baselines:
- Motivation:
- We're targeting Python-3.5 and above.
- Ubuntu LTS (16.04) already ships without python2. RHEL8 is coming which is also expected to be python3 based. These considerations make this high priority.
- Ansible users are getting restless: https://groups.google.com/forum/#!topic/ansible-project/DUKzTho3OCI
- Goals for 2.2:
- This is probably going to take multiple releases to complete; need to get started now
- Tech preview level of support
- Baselines:
- Controller-side code can run on Python3
- We're targeting Python-3.5 and above.
- Update: Essential features have been shown to work on Python3.
Currently all unittests and all but three integration tests are
- Goals for 2.2:
passing on Python3. Code has not been line-by-line audited so bugs
remain but can be treated as bugs, not as massive, invasive new features.
- Tech preview level of support
- Almost all of our deps have been ported:
- Controller-side code can run on Python3
- The base deps in setup.py are ported: ['paramiko', 'jinja2', "PyYAML", 'setuptools', 'pycrypto >= 2.6']
- Update: Essential features have been shown to work on Python3.
- python-six from the rpm spec file has been ported
Currently all unittests and all but three integration tests are
- Python-keyczar from the rpm spec file is not.
passing on Python3. Code has not been line-by-line audited so bugs
- Strategy: removing keyczar when we drop accelerate for 2.3. Print deprecation in 2.1.
remain but can be treated as bugs, not as massive, invasive new features.
- Almost all of our deps have been ported:
- Module_utils ported to dual python3/python2(2.4 for much of it, python2.6 for specific things)
**Mostly done:** Also not line-by-line audited but the unittests
- The base deps in setup.py are ported: ['paramiko', 'jinja2', "PyYAML", 'setuptools', 'pycrypto >= 2.6']
and integration tests do show that the most use functionality is working.
- python-six from the rpm spec file has been ported
- Add module_utils files to help port
- Python-keyczar from the rpm spec file is not.
- Update: copy of the six library (v1.4.1 for python2.4 compat) and unicode helpers are here (ansible.module_utils._text.{to_bytes,to_text,to_native})
- Strategy: removing keyczar when we drop accelerate for 2.3. Print deprecation in 2.1.
- A few basic modules ported to python3
- Stat module best example module since it’s essential.
- Module_utils ported to dual python3/python2(2.4 for much of it, python2.6 for specific things)
- Update:
**Mostly done:** Also not line-by-line audited but the unittests
and integration tests do show that the most use functionality is working.
- A handful of modules like stat have been line-by-line ported. They should work reliably with few python3-specific bugs. All but three integration tests pass which means that most essential modules are working to some extent on Python3.
- Add module_utils files to help port
- The three failing tests are: service, hg, and uri.
- Note, large swaths of the modules are not tested. The status of
- Update: copy of the six library (v1.4.1 for python2.4 compat) and unicode helpers are here (ansible.module_utils._text.{to_bytes,to_text,to_native})
- A few basic modules ported to python3
- Stat module best example module since it’s essential.
- Update:
- A handful of modules like stat have been line-by-line ported. They should work reliably with few python3-specific bugs. All but three integration tests pass which means that most essential modules are working to some extent on Python3.
- The three failing tests are: service, hg, and uri.
- Note, large swaths of the modules are not tested. The status of
these is unknown
these is unknown
- All code should compile under Python3.
- lib/ansible/* and all modules now compile under Python-3.5
- Side work to do:
- All code should compile under Python3.
- Figure out best ways to run unit-tests on modules. Start unit-testing modules. This is going to become important so we don’t regress python3 or python2.4 support in modules (Going to largely punt on this for 2.2. Matt Clay is working on building us a testing foundation for the first half of 2.2 development so we’ll re-evaluate towards the middle of the dev cycle).
- lib/ansible/* and all modules now compile under Python-3.5
- More unit tests of module_utils
- More integration tests. Currently integration tests are the best way to test ansible modules so we have to rely on those.
- Side work to do:
- Figure out best ways to run unit-tests on modules. Start unit-testing modules. This is going to become important so we don’t regress python3 or python2.4 support in modules (Going to largely punt on this for 2.2. Matt Clay is working on building us a testing foundation for the first half of 2.2 development so we’ll re-evaluate towards the middle of the dev cycle).
- More unit tests of module_utils
- More integration tests. Currently integration tests are the best way to test ansible modules so we have to rely on those.
- Goals for 2.3:
- Goals for 2.3:
- Bugfixing, bugfixing, bugfixing. We need community members to test,
submit bugs, and add new unit and integration tests. I'll have some
time allocated both to review any Python3 bugfixes that they submit
and to work on bug reports without PRs. The overall goal is to make
the things that people do in production with Ansible work on Python 3.
- **Infrastructure Buildout and Changes** (Matt Clay)
- Bugfixing, bugfixing, bugfixing. We need community members to test,
Another note from Jason M: A lot of this work is to ease the burden of CI, CI performance, increase our testing coverage and all of that sort of thing. It’s not necessarily feature work, but it’s \*\*critical\*\* to growing our product and our ability to get community changes in more securely and quickly.
submit bugs, and add new unit and integration tests. I'll have some
time allocated both to review any Python3 bugfixes that they submit
and to work on bug reports without PRs. The overall goal is to make
the things that people do in production with Ansible work on Python 3.
Infrastructure Buildout and Changes
-----------------------------------
Lead by Matt Clay
Another note from Jason M: A lot of this work is to ease the burden of CI, CI performance, increase our testing coverage and all of that sort of thing. It’s not necessarily feature work, but it’s \*\*critical\*\* to growing our product and our ability to get community changes in more securely and quickly.
- **CI Performance**
- **CI Performance**
Reduce time spent waiting on CI for PRs. Combination of optimizing existing Travis setup and offloading work to other services. Will be impacted by available budget.
Reduce time spent waiting on CI for PRs. Combination of optimizing existing Travis setup and offloading work to other services. Will be impacted by available budget.
**Done:** Most tests have been migrated from Travis to Shippable.
**Done:** Most tests have been migrated from Travis to Shippable.
- **Core Module Test Organization**
- **Core Module Test Organization**
Relocate core module tests to ansible-modules-core to encourage inclusion of tests in core module PRs.
Relocate core module tests to ansible-modules-core to encourage inclusion of tests in core module PRs.
**Deferred:** Relocation of core module tests has been deferred due to proposed changes in `modules management <https://github.com/ansible/proposals/blob/master/modules-management.md>`_.
**Deferred:** Relocation of core module tests has been deferred due to proposed changes in `modules management <https://github.com/ansible/proposals/blob/master/modules-management.md>`_.
- **Documentation**
- **Documentation**
Expand documentation on setting up a development and test environment, as well as writing tests. The goal is to ease development for new contributors and encourage more testing, particularly with module contributions.
Expand documentation on setting up a development and test environment, as well as writing tests. The goal is to ease development for new contributors and encourage more testing, particularly with module contributions.
- **Test Coverage**
- **Test Coverage**
- Expand test coverage, particularly for CI. Being testing, this is open ended. Will be impacted by available budget.
- Expand test coverage, particularly for CI. Being testing, this is open ended. Will be impacted by available budget.
**Done:** Module PRs now run integration tests for the module(s) being changed.
**Done:** Module PRs now run integration tests for the module(s) being changed.
- Python 3 - Run integration tests using Python 3 on CI with tagging for those which should pass, so we can track progress and detect regressions.
- Python 3 - Run integration tests using Python 3 on CI with tagging for those which should pass, so we can track progress and detect regressions.
**Done:** Integration tests now run on Shippable using a Ubuntu 16.04 docker image with only Python 3 installed.
**Done:** Integration tests now run on Shippable using a Ubuntu 16.04 docker image with only Python 3 installed.
- Windows - Create framework for running Windows integration tests, ideally both locally and on CI.
- Windows - Create framework for running Windows integration tests, ideally both locally and on CI.
**Done:** Windows integration tests now run on Shippable.
**Done:** Windows integration tests now run on Shippable.
- FreeBSD - Include FreeBSD in CI coverage. Not originally on the roadmap, this is an intermediary step for CI coverage for OS X.
- FreeBSD - Include FreeBSD in CI coverage. Not originally on the roadmap, this is an intermediary step for CI coverage for OS X.
**Done:** FreeBSD integration tests now run on Shippable.
**Done:** FreeBSD integration tests now run on Shippable.
- OS X - Include OS X in CI coverage.
- OS X - Include OS X in CI coverage.
**Done:** OS X integration tests now run on Shippable.
**Done:** OS X integration tests now run on Shippable.
- The 2.3 Ansible Core is just a little different than the past two major releases we've done. In addition to feature work, we're using part of the time for this release to reduce some of our backlog in other areas than pure development.
- The 2.3 Ansible Core is just a little different than the past two major releases we've done. In addition to feature work, we're using part of the time for this release to reduce some of our backlog in other areas than pure development.
- *Administration:* Clean up our GitHub repos and move to one repo so that contributions, tickets, submissions, etc are centralized and easier for both the community and the Core Team to manage.
- *Administration:* Clean up our GitHub repos and move to one repo so that contributions, tickets, submissions, etc are centralized and easier for both the community and the Core Team to manage.
- *Metadata:* Move to a Metadata based system for modules. This has been discussed here: https://github.com/ansible/proposals/blob/master/modules-management.md
- *Metadata:* Move to a Metadata based system for modules. This has been discussed here: https://github.com/ansible/proposals/blob/master/modules-management.md
- *Documentation:* We're aware that Docs have issues. Scott Butler, aka Dharmabumstead will be leading the charge on how he and we as a community can clean them up.
- *Documentation:* We're aware that Docs have issues. Scott Butler, aka Dharmabumstead will be leading the charge on how he and we as a community can clean them up.
- *Backlog & Stability:* We're spending some of the cycles for 2.3 trying to reduce our ticket/PR backlog, and clean up some particular areas of the project that the community has expressed particular frustrations about.
- *Backlog & Stability:* We're spending some of the cycles for 2.3 trying to reduce our ticket/PR backlog, and clean up some particular areas of the project that the community has expressed particular frustrations about.
- *Python 3:* The community and Toshio have done TONS of work getting Python 3 working. Still more to go...
- *Python 3:* The community and Toshio have done TONS of work getting Python 3 working. Still more to go...
- *Features:* We still have some cool stuff coming. Check it out below. For people on the Networking side of the world, the Persistent Connection Manager will be a *huge* feature and performance gain.
- *Features:* We still have some cool stuff coming. Check it out below. For people on the Networking side of the world, the Persistent Connection Manager will be a *huge* feature and performance gain.
Repo Merge
----------
- **Repo Merge**
- Script that a submitter can run to migrate their PR **(done)**
- Script that a committer can run to fork a PR and then merge to ansible/ansible **(mostly done)**
- Script that a submitter can run to migrate their PR **(done)**
- Move all the issues (remove old ones that can be removed) **(done)**
- Script that a committer can run to fork a PR and then merge to ansible/ansible **(mostly done)**
- Enhance ansibullbot to accommodate the changes (jctanner) **(in progress, going well)**
- Move all the issues (remove old ones that can be removed) **(done)**
- Enhance ansibullbot to accommodate the changes (jctanner) **(in progress, going well)**
Metadata
--------
- **Metadata**
- Add metadata to the modules we ship **(done)**
- Write code to use metadata in docs **(done)**
- Add metadata to the modules we ship **(done)**
- If needed for python2/3 write code to use metadata in module_common or pluginloader **(not needed)**
- Write code to use metadata in docs **(done)**
- If needed for python2/3 write code to use metadata in module_common or pluginloader **(not needed)**
Documentation
-------------
- **Documentation**
- Update developing_modules **(in progress, will continue in 2.4)**
- Set up rst skeleton for module_utils docs.
- Update developing_modules **(in progress, will continue in 2.4)**
- Plugin development docs
- Set up rst skeleton for module_utils docs.
- Speed up `make webdocs` https://github.com/ansible/ansible/issues/17406 **(done)**
- Plugin development docs
- Speed up `make webdocs` https://github.com/ansible/ansible/issues/17406 **(done)**
Windows
-------
- **Windows platform** (nitzmahone)
Lead by nitzmahone
- **Platform**
- Pipelining support **(done)**
- Pipelining support **(done)**
- Become support **(done/experimental)**
- Become support **(done/experimental)**
@ -48,7 +50,7 @@ Target: Mid April 2017
- Kerberos encryption (via notting, pywinrm/requests_kerberos/pykerberos) **(in progress, available in pywinrm post 2.3 release)**
- Kerberos encryption (via notting, pywinrm/requests_kerberos/pykerberos) **(in progress, available in pywinrm post 2.3 release)**
- Fix plugin-specific connection var lookup/delegation (either registered explicitly by plugins or ansible_(plugin)_*) **(bumped to 2.4)**
- Fix plugin-specific connection var lookup/delegation (either registered explicitly by plugins or ansible_(plugin)_*) **(bumped to 2.4)**
- **Windows modules** (nitzmahone)
- **Modules**
- win_domain module **(done)**
- win_domain module **(done)**
- win_domain_membership module **(done)**
- win_domain_membership module **(done)**
@ -61,93 +63,102 @@ Target: Mid April 2017
- Updates to win_updates, adopt to core (stretch) **(bump to 2.4)**
- Updates to win_updates, adopt to core (stretch) **(bump to 2.4)**
- Updates to win_package, adopt to core (+ deprecate win_msi) (stretch) **(bump to 2.4)**
- Updates to win_package, adopt to core (+ deprecate win_msi) (stretch) **(bump to 2.4)**
- **Azure modules** (nitzmahone/mattclay)
Azure
-----
- Ensure Azure SDK rc6/RTM work **(done)**
Lead by nitzmahone, mattclay
- Move tests from ansible/azure_rm repo to ansible/ansible **(bump to 2.4, no CI resources)**
- Update/enhance tests **(bump to 2.4, no CI resources)**
- Just the mercurial tests left because we haven't created an image with
- Just the mercurial tests left because we haven't created an image with
both python2 and python3 to test it on yet.
both python2 and python3 to test it on yet.
- Check by doing ``grep skip/python3 test/integration/targets/*/aliases``
- Check by doing ``grep skip/python3 test/integration/targets/*/aliases``
- If users report bugs on python3, these should be fixed and will prioritize our work on porting other modules.
- Still have to solve the python3-only and python2-only modules. Thinking of doing this via metadata. Will mean we have to use metadata at the module_common level. Will also mean we don’t support py2-only or py3-only old style python modules.
- Note: Most of the currently tested ansible features now run. But there’s still a lot of code that’s untested.
- **Testing and CI** (mattclay)
- If users report bugs on python3, these should be fixed and will prioritize our work on porting other modules.
- *Static Code Analysis:* Create custom pylint extensions to automate detection of common Ansible specific issues reported during code review. Automate feedback on PRs for new code only to avoid noise from existing code which does not pass.
- Still have to solve the python3-only and python2-only modules. Thinking of doing this via metadata. Will mean we have to use metadata at the module_common level. Will also mean we don’t support py2-only or py3-only old style python modules.
- Note: Most of the currently tested ansible features now run. But there’s still a lot of code that’s untested.
**Ongoing:** Some static code analysis is now part of the CI process:
Testing and CI
--------------
Lead by mattclay
- pep8 is now being run by CI, although not all PEP 8 rules are being enforced.
- *Static Code Analysis:* Create custom pylint extensions to automate detection of common Ansible specific issues reported during code review. Automate feedback on PRs for new code only to avoid noise from existing code which does not pass.
- pylint is now being run by CI, but currently only on the ansible-test portion of codebase.
- *Test Reliability:* Eliminate transient test failures by fixing unreliable tests. Reduce network dependencies by moving network resources into httptester.
**Ongoing:** Some static code analysis is now part of the CI process:
**Ongoing:** Many of the frequent sources of test instability have been resolved. However, more work still remains.
- pep8 is now being run by CI, although not all PEP 8 rules are being enforced.
- pylint is now being run by CI, but currently only on the ansible-test portion of codebase.
Some new issues have also appeared, which are currently being worked on.
- *Test Reliability:* Eliminate transient test failures by fixing unreliable tests. Reduce network dependencies by moving network resources into httptester.
- *Enable Remaining Tests:* Implement fixes for OS X, FreeBSD and Python 3 to enable the remaining blacklisted tests for CI.
**Ongoing:** Many of the frequent sources of test instability have been resolved. However, more work still remains.
**Ongoing:** More tests have been enabled for OS X, FreeBSD and Python 3. However, work still remains to enable more tests.
Some new issues have also appeared, which are currently being worked on.
- *Windows Server 2016:* Add Windows Server 2016 to CI when official AMIs become available.
- *Enable Remaining Tests:* Implement fixes for OS X, FreeBSD and Python 3 to enable the remaining blacklisted tests for CI.
**Delayed:** Integration tests pass on Windows Server 2016. However, due to intermittent WinRM issues, the tests have been disabled.
**Ongoing:** More tests have been enabled for OS X, FreeBSD and Python 3. However, work still remains to enable more tests.
Once the issues with WinRM have been resolved, the tests will be re-enabled.
- *Windows Server 2016:* Add Windows Server 2016 to CI when official AMIs become available.
- *Repository Consolidation:* Update CI to maintain and improve upon existing functionality after repository consolidation.
**Delayed:** Integration tests pass on Windows Server 2016. However, due to intermittent WinRM issues, the tests have been disabled.
**Done:** A new test runner, ansible-test, has been deployed to manage CI jobs on Shippable.
Once the issues with WinRM have been resolved, the tests will be re-enabled.
Tests executed on PRs are based on the changes made in the PR, for example:
- *Repository Consolidation:* Update CI to maintain and improve upon existing functionality after repository consolidation.
- Changes to a module will only run tests appropriate for that module.
**Done:** A new test runner, ansible-test, has been deployed to manage CI jobs on Shippable.
- Changes to Windows modules or the Windows connection plugin run tests on Windows.
- Changes to network modules run tests on the appropriate virtual network device (currently supporting VyOS and IOS).
Tests executed on merges are based on changes since the last successful merge test.
Tests executed on PRs are based on the changes made in the PR, for example:
- **Amazon resources** (ryansb)
- Changes to a module will only run tests appropriate for that module.
- Changes to Windows modules or the Windows connection plugin run tests on Windows.
- Changes to network modules run tests on the appropriate virtual network device (currently supporting VyOS and IOS).
- Improve ec2.py integration tests **(partial, more to do in 2.4)**
Tests executed on merges are based on changes since the last successful merge test.
- ELB version 2 **(pushed - needs_revision [PR](https://github.com/ansible/ansible/pull/19491))**
- CloudFormation YAML, cross-stack reference, and roles support **(done)**
- Starting with 2.4, all items that are deprecated will be removed in 4 major releases unless otherwise stated.
-------------------------
- Starting with 2.4, all items that are deprecated will be removed in 4 major releases unless otherwise stated.
- For example: A module that is deprecated in 2.4 will be removed in 2.8
- For example: A module that is deprecated in 2.4 will be removed in 2.8
- **Python 2.4 and 2.5 support discontinuation**
Python 2.4 and 2.5 support discontinuation
- Ansible will not support Python 2.4 nor 2.5 on the target hosts anymore. Going forward, Python 2.6+ will be required on targets, as already is the case on the controller.
------------------------------------------
- Ansible will not support Python 2.4 nor 2.5 on the target hosts anymore.
- **Python 3 and beyond**
Going forward, Python 2.6+ will be required on targets, as already is the case on the controller.
- Ansible Core Engine and Core modules will be tested on Python 3
Python 3
- Communicate with Linux distros to provide Ansible running on Python 3
--------
- Check for Python 3 tests on core modules and create any missing
- Ansible Core Engine and Core modules will be tested on Python 3
- Communicate with Linux distros to provide Ansible running on Python 3
- **Ansible-Config**
- Check for Python 3 tests on core modules and create any missing
- New yaml format for config
Ansible-Config
- Extend the ability of the current config system by adding an ansible-config command and add the following:
--------------
- New yaml format for config
- Dump existing config settings
- Extend the ability of the current config system by adding an ``ansible-config`` command and add the following:
- Update / write a config entry
- Dump existing config settings
- Update / write a config entry
- Show available options (ini entry, yaml, env var, etc)
- Show available options (ini entry, yaml, env var, etc)
- Proposal found in ansible/proposals issue `#35 <https://github.com/ansible/proposals/issues/35>`_.
- Proposal found in ansible/proposals issue `#35 <https://github.com/ansible/proposals/issues/35>`_.
- Initial PR of code found in ansible/ansible PR `#12797 <https://github.com/ansible/ansible/pull/12797>`_.
- Initial PR of code found in ansible/ansible PR `#12797 <https://github.com/ansible/ansible/pull/12797>`_.
- **Inventory Overhaul**
Inventory
---------
- Current inventory is overly complex, non modular and mostly still a legacy from inception.
- Current inventory is overly complex, non modular and mostly still a legacy from inception.
- We also want to add a common set of features to most inventory sources but are hampered by the current code base.
- We also want to add a common set of features to most inventory sources but are hampered by the current code base.
- Proposal found in ansible/proposals issue `#41 <https://github.com/ansible/proposals/issues/41>`_.
- Proposal found in ansible/proposals issue `#41 <https://github.com/ansible/proposals/issues/41>`_.
- **Facts Refreshening**
Facts
-----
- Configurable list of ‘fact modules’ for ``gather_facts``
- Configurable list of ‘fact modules’ for ``gather_facts``
- Fact gathering policy finer grained
- Fact gathering policy finer grained
- Make ``setup.py``/``facts`` more pluggable
- Make ``setup.py``/``facts`` more pluggable
- Improve testing of ``setup.py``/``facts.py``
- Improve testing of ``setup.py``/``facts.py``
- Namespacing fact variables (via a config option) implemented in ansible/ansible PR `#18445 <https://github.com/ansible/ansible/pull/18445>`_.
- Namespacing fact variables (via a config option) implemented in ansible/ansible PR `#18445 <https://github.com/ansible/ansible/pull/18445>`_.
- Proposal found in ansible/proposals issue `#17 <https://github.com/ansible/proposals/issues/17>`_.
Proposal found in ansible/proposals issue `#17 <https://github.com/ansible/proposals/issues/17>`_.
- **PluginLoader Refactor**
PluginLoader
------------
- Over the past couple releases we've had some thoughts about how
- Over the past couple releases we've had some thoughts about how
PluginLoader might be better structured
PluginLoader might be better structured
- Load the loaders via an initialization function(), not when importing
- Load the loaders via an initialization function(), not when importing
the module. (stretch goal, doesn't impact the CLI)
the module. (stretch goal, doesn't impact the CLI)
- Separate duties of ``PluginLoader`` from ``PluginFinder``. Most plugins need
- Separate duties of ``PluginLoader`` from ``PluginFinder``. Most plugins need
both but Modules and Module_utils only need a PluginFinder
both but Modules and Module_utils only need a PluginFinder
- Write different ``PluginFinder`` subclasses for module_utils and perhaps
- Write different ``PluginFinder`` subclasses for module_utils and perhaps
Modules. Most Plugin types have a flattened namespace and are single
Modules. Most Plugin types have a flattened namespace and are single
python files. Modules include code that is not written in python.
python files. Modules include code that is not written in python.
Module_utils are vastly different from the other Plugins as they
Module_utils are vastly different from the other Plugins as they
maintain a hierarchical namespace and are multi-file.
maintain a hierarchical namespace and are multi-file.
- Potentially split module_utils loader for python from module_utils
- Potentially split module_utils loader for python from module_utils
loader for powershell. Currently we only support generic module_utils
loader for powershell. Currently we only support generic module_utils
for python modules. The powershell modules always include a single,
for python modules. The powershell modules always include a single,
hardcoded powershell module_utils file. If we add generic module_utils
hardcoded powershell module_utils file. If we add generic module_utils
for powershell, we'll need to decide how to organize the code.
for powershell, we'll need to decide how to organize the code.
- **Static Loop Keyword**
Static Loop Keyword
-------------------
- Deprecate (not on standard deprecation cycle) ``with_`` in favor of ``loop:``
- Deprecate (not on standard deprecation cycle) ``with_`` in favor of ``loop:``
- This ``loop:`` will take only a list
- This ``loop:`` will take only a list
- Remove complexity from loops, lookups are still available to users
- Remove complexity from loops, lookups are still available to users
- Less confusing having a static directive vs a one that is dynamic depending on plugins loaded.
- Less confusing having a static directive vs a one that is dynamic depending on plugins loaded.
- **Vault Extensibility**
Vault
-----
- Support for multiple vault passwords
- Support for multiple vault passwords
- Each decrypted item should know which secret to request
- Each decrypted item should know which secret to request
- Support requesting credentials (password prompt) as callbacks
- Support requesting credentials (password prompt) as callbacks
- Ability to open and edit file with encrypted vars deencrypted, and encrypt/format on save
- Ability to open and edit file with encrypted vars deencrypted, and encrypt/format on save
- **Globalize Callbacks**
Globalize Callbacks
-------------------
- Make send_callback available to other code that cannot use it.
- Make send_callback available to other code that cannot use it.
- Would allow for ‘full formatting’ of output (see JSON callback)
- Would allow for ‘full formatting’ of output (see JSON callback)
- Fixes static ‘include’ display problem
- Fixes static ‘include’ display problem
- **Document Plugins**
Plugins
-------
- Allow plugins to have embedded docs (like modules)
- Allow plugins to have embedded docs (like modules)
- Update ansible-doc and website to generate docs from these ansible/ansible PR `#22796 <https://github.com/ansible/ansible/pull/22796>`_.
- Update ansible-doc and website to generate docs from these ansible/ansible PR `#22796 <https://github.com/ansible/ansible/pull/22796>`_.
- **Group Priorities**
Group Priorities
----------------
- Start using existing group priority variable to sort/merge group vars
- Start using existing group priority variable to sort/merge group vars
- Implementation for this in ansible/ansible PR `#22580 <https://github.com/ansible/ansible/pull/22580>`_.
- Implementation for this in ansible/ansible PR `#22580 <https://github.com/ansible/ansible/pull/22580>`_.
- Documentation of group priority variable
- Documentation of group priority variable
- **Runtime Check on Modules for Blacklisting**
Runtime Check on Modules for Blacklisting
-----------------------------------------
- Filter on things like "supported_by" in module metadata
- Filter on things like "supported_by" in module metadata
- Provide users with an option of "warning, error or allow/ignore"
- Provide users with an option of "warning, error or allow/ignore"
- Configurable via ansible.cfg and environment variable
- Configurable via ansible.cfg and environment variable
- **Disambiguate Includes**
Disambiguate Includes
---------------------
- Create import_x for ‘static includes’ (import_task, import_play, import_role)
- Create import_x for ‘static includes’ (import_task, import_play, import_role)
- Any directives are applied to the ‘imported’ tasks
- Any directives are applied to the ‘imported’ tasks
- Create include_x for ‘dynamic includes’ (include_task, include_role)
- Create include_x for ‘dynamic includes’ (include_task, include_role)
- Any directives apply to the ‘include’ itself
- Any directives apply to the ‘include’ itself
- **Windows Support**
Windows
-------
- New PS/.NET module API
- New PS/.NET module API
- Windows Nano Server support
- Windows Nano Server support
- Windows module_utils pluginloader
- Windows module_utils pluginloader
- Refactor duplicated module code into new module_utils files
- Refactor duplicated module code into new module_utils files
- Evaluate #Requires directives (existing and new: PS version, OS version, etc)
- Evaluate #Requires directives (existing and new: PS version, OS version, etc)
- Improve module debug support/persistence
- Improve module debug support/persistence
- Explore official DSC support
- Explore official DSC support
- Explore module intermediate output
- Explore module intermediate output
- Explore Powershell module unit testing
- Explore Powershell module unit testing
- Explore JEA support (stretch)
- Explore JEA support (stretch)
- Extended become support with network/service/batch logon types
- Extended become support with network/service/batch logon types
- Module updates
- Module updates
- Split "Windows" category into multiple subs
- Split "Windows" category into multiple subs
- Domain user/group management modules
- Domain user/group management modules
- win_mapped_drive module
- win_mapped_drive module
- win_hotfix
- win_hotfix
- win_updates rewrite to require become
- win_updates rewrite to require become
- win_package changes required to deprecate win_msi
- win_package changes required to deprecate win_msi
- win_copy re-write
- win_copy re-write
- **Cloud Provider Support**
AWS
---
- AWS
- Focus on pull requests for various modules
- Triage existing merges for modules
- Focus on pull requests for various modules
- Module work
- Triage existing merges for modules
- Module work
- elb-target-groups
- alb*
- elb-target-groups
- ecs
- alb*
- Data Pipelines
- ecs
- VPN
- Data Pipelines
- DirectConnect
- VPN
- DirectConnect
Azure
-----
- Azure
- Expose endpoint overrides
- Reformat/document module output to collapse internal API structures and surface important data (eg, public IPs, NICs, data disks)
- Expose endpoint overrides
- Add load balancer module
- Reformat/document module output to collapse internal API structures and surface important data (eg, public IPs, NICs, data disks)
- Add Azure Functions module
- Add load balancer module
- Add Azure Functions module
Google Cloud Platform
---------------------
- Google Cloud Platform
- New Module: DataProc
- Support for Cross-Region HTTP Load Balancing
- New Module: DataProc
- New Module: GKE
- Support for Cross-Region HTTP Load Balancing
- New Module: GKE
Network Roadmap
---------------
- **Network Roadmap**
- Removal of ``*_template`` modules
- Session Tracing
- Removal of ``*_template`` modules
- Refactor ansible-connection to cli
- Session Tracing
- Module Work
- Refactor ansible-connection to cli
- Module Work
- Declarative intent modules
- OpenVSwitch
- Declarative intent modules
- OpenVSwitch
Contributor Quality of Life
---------------------------
- **Contributor Quality of Life**
- All Core and Curated modules will work towards having unit testing.
- More bot improvements!
- All Core and Curated modules will work towards having unit testing.
- Test Infrastructure changes
- More bot improvements!
- Test Infrastructure changes
- Shippable + Bot Integration
- Shippable + Bot Integration
- Provide verified test results to the bot from Shippable so the bot can comment on PRs with CI failures.
- Enable the bot to mark PRs with ``ci_verified`` if all CI failures are verified.
- Provide verified test results to the bot from Shippable so the bot can comment on PRs with CI failures.
- Enable the bot to mark PRs with ``ci_verified`` if all CI failures are verified.
- Windows Server 2016 Integration Tests
- Windows Server 2016 Integration Tests
- Restore Windows Server 2016 integration tests on Shippable.
- Restore Windows Server 2016 integration tests on Shippable.
- Originally enabled during the 2.3 release cycle, but later disabled due to intermittent WinRM issues.
- Depends on resolution of WinRM connection issues.
- Originally enabled during the 2.3 release cycle, but later disabled due to intermittent WinRM issues.
- Depends on resolution of WinRM connection issues.
- Windows Server Nano Integration Tests
- Windows Server Nano Integration Tests
- Add support to ansible-core-ci for Windows Server 2016 Nano and enable on Shippable.
- Add support to ansible-core-ci for Windows Server 2016 Nano and enable on Shippable.
- This will use a subset of the existing Windows integration tests.
- This will use a subset of the existing Windows integration tests.
- Depends on resolution of WinRM connection issues.
- Depends on resolution of WinRM connection issues.
- Windows + Python 3 Tests
- Windows + Python 3 Tests
- Run basic Windows tests using Python 3 as the controller.
- Run basic Windows tests using Python 3 as the controller.
- Depends on resolution of WinRM Python 3 issues.
- Depends on resolution of WinRM Python 3 issues.
- Cloud Integration Tests
- Cloud Integration Tests
- Run existing cloud integration tests for AWS, Azure and GCP as part of CI.
- Run existing cloud integration tests for AWS, Azure and GCP as part of CI.
- Tests to be run only on cloud module (and module_utils) PRs and merges for the relevant cloud provider.
- Tests to be run only on cloud module (and module_utils) PRs and merges for the relevant cloud provider.
- Test Reliability
- Test Reliability
- Further improve test reliability to reduce false positives on Shippable.
- Further improve test reliability to reduce false positives on Shippable.
- This continues work from the 2.3 release cycle.
- This continues work from the 2.3 release cycle.
- Static Code Analysis
- Static Code Analysis
- Further expand the scope and coverage of static analysis.
- Further expand the scope and coverage of static analysis.
Dag Wieers
7 years ago
committed by