ironic/doc/source/dev/code-contribution-guide.rst

Code Contribution Guide

This document provides some necessary points for developers to consider when writing and reviewing Ironic code. The checklist will help developers get things right.

Adding New Features

Starting with the Mitaka development cycle, Ironic tracks new features using RFEs (Requests for Feature Enhancements) instead of blueprints. These are bugs with 'rfe' tag, and they should be submitted before a spec or code is proposed. When a member of ironic-drivers launchpad team decides that the proposal is worth implementing, a spec (if needed) and code should be submitted, referencing the RFE bug. Contributors are welcome to submit a spec and/or code before the RFE is approved, however those patches will not land until the RFE is approved.

Here is a list of steps to do during the new process of adding a new feature to Ironic:

1. Submit a bug report at https://bugs.launchpad.net/ironic/+filebug. There are two fields that must be filled: 'summary' and 'further information'. The 'summary' must be brief enough to fit in one line: if you can’t describe it in a few words it may mean that you are either trying to capture more than one RFE at once, or that you are having a hard time defining what you are trying to solve at all.
2. Describe the proposed change in the 'further information' field. The description should provide enough details for a knowledgeable developer to understand what is the existing problem in the current platform that needs to be addressed, or what is the enhancement that would make the platform more capable, both from a functional and a non-functional standpoint.
3. Submit the bug, add an 'rfe' tag to it and assign yourself or whoever is going to work on this feature.
4. As soon as a member of the ironic-drivers team acknowledges the bug, it will be moved into the 'Triaged' state. The importance will be set to 'Wishlist' to signal the fact that the report is indeed a feature and there is no severity associated to it. Discussion about the RFE, and whether to approve it, happens in bug comments while in the 'Triaged' state.
5. The ironic-drivers team will evaluate the RFE and may advise the submitter to file a spec in ironic-specs to elaborate on the feature request, in case the RFE requires extra scrutiny, more design discussion, etc. For the spec submission process, please see the Ironic Specs Process.
6. If a spec is not required, once the discussion has happened and there is positive consensus among the ironic-drivers team on the RFE, the RFE is 'approved', and its tag will move from 'rfe' to 'rfe-approved'. This means that the feature is approved and the related code may be merged.
7. If a spec is required, the spec must be submitted (with the bug properly referenced as 'Partial-Bug' in the commit message), reviewed, and merged before the RFE will be 'approved' (and the tag changed to 'rfe-approved').
8. The bug then goes through the usual process -- first to 'In progress' when the spec/code is being worked on, then 'Fix Released' when it is implemented.
9. If the RFE is rejected, the ironic-drivers team will move the bug to "Won't Fix" status.

When working on an RFE, please be sure to tag your commits properly: "Partial-Bug: #xxxx" or "Related-Bug: #xxxx" for intermediate commits for the feature, and "Closes-Bug: #xxxx" for the final commit. It is also helpful to set a consistent review topic, such as "bug/xxxx" for all patches related to the RFE.

If the RFE spans across several projects (e.g. ironic and python-ironicclient), but the main work is going to happen within ironic, please use the same bug for all the code you're submitting, there is no need to create a separate RFE in every project.

Note that currently the Ironic bug tracker is managed by the open 'ironic-bugs' team, not the ironic-drivers team. This means that anyone may edit bug details, and there is room to game the system here. RFEs may only be approved by members of the ironic-drivers team. Attempts to sneak around this rule will not be tolerated, and will be called out in public on the mailing list.

Live Upgrade Related Concerns

Ironic implements upgrade with the same methodology of Nova:

http://docs.openstack.org/developer/nova/upgrade.html

Ironic API RPC Versions

• When the signature(arguments) of an RPC method is changed, the following things need to be considered:

• The RPC version must be incremented and be the same value for both the client (conductor/rpcapi.py, used by ironic-api) and the server (conductor/manager.py, used by ironic-conductor).
• New arguments of the method can only be added as optional. Existing arguments cannot be removed or changed in incompatible ways (with the method in older RPC versions).
• Client-side can pin a version cap by passing version_cap to the constructor of oslo_messaging.RPCClient. Methods which change arguments should run client.can_send_version() to see if the version of the request is compatible with the version cap of RPC Client, otherwise the request needs to be created to work with a previous version that is supported.
• Server-side should tolerate the older version of requests in order to keep working during the progress of live upgrade. The behavior of server-side should depend on the input parameters passed from the client-side.

Object Versions

• When Object classes (subclasses of ironic.objects.base.IronicObject) are modified, the following things need to be considered:

• The change of fields and the signature of remotable method needs a bump of object version.
• The arguments of methods can only be added as optional, they cannot be removed or changed in an incompatible way.
• Fields types cannot be changed. If it is a must, create a new field and deprecate the old one.
• When new version objects communicate with old version objects, obj_make_compatible() will be called to convert objects to the target version during serialization. So objects should implement their own obj_make_compatible() to remove/alter attributes which was added/changed after the target version.
• There is a test (object/test_objects.py) to generate the hash of object fields and the signatures of remotable methods, which helps developers to check if the change of objects need a version bump. The object fingerprint should only be updated with a version bump.

Driver Internal Info

The driver_internal_info node field was introduced in the Kilo release. It allows driver developers to store internal information that can not be modified by end users. Here is the list of existing common and agent driver attributes:

Common attributes:

• is_whole_disk_image: A Boolean value to indicate whether the user image contains ramdisk/kernel.
• clean_steps: An ordered list of clean steps that will be performed on the node.
• instance: A list of dictionaries containing the disk layout values.
• root_uuid_or_disk_id: A String value of the bare metal node's root partition uuid or disk id.
• persistent_boot_device: A String value of device from ironic.common.boot_devices.
• is_next_boot_persistent: A Boolean value to indicate whether the next boot device is persistent_boot_device.

Agent driver attributes:

• agent_url: A String value of IPA API URL so that Ironic can talk to IPA ramdisk.
• agent_last_heartbeat: An Integer value of the last agent heartbeat time.
• hardware_manager_version: A String value of the version of the hardware manager in IPA ramdisk.
• target_raid_config: A Dictionary containing the target RAID configuration. This is a copy of the same name attribute in Node object. But this one is never actually saved into DB and is only read by IPA ramdisk.

Note

These are only some fields in use. Other vendor drivers might expose more driver_internal_info properties, please check their development documentation and/or module docstring for details. It is important for developers to make sure these properties follow the precedent of prefixing their variable names with a specific interface name (e.g., iboot_bar, amt_xyz), so as to minimize or avoid any conflicts between interfaces.

Ironic Specs Process

Specifications must follow the template which can be found at specs/template.rst <http://git.openstack.org/cgit/openstack/ironic-specs/tree/ specs/template.rst>, which is quite self-documenting. Specifications are proposed by adding them to the specs/approved directory, adding a soft link to it from the specs/not-implemented directory, and posting it for review to Gerrit. For more information, please see the README <http://git. openstack.org/cgit/openstack/ironic-specs/tree/README.rst>.

The same Gerrit process as with source code, using the repository ironic-specs <http://git.openstack.org/cgit/openstack/ ironic-specs/>, is used to add new specifications.

All approved specifications are available at: http://specs.openstack.org/openstack/ironic-specs. If a specification has been approved but not completed within one or more releases since the approval, it may be re-reviewed to make sure it still makes sense as written.

Ironic specifications are part of the RFE (Requests for Feature Enhancements) process. You are welcome to submit patches associated with an RFE, but they will have a -2 ("do not merge") until the specification has been approved. This is to ensure that the patches don't get accidentally merged beforehand. You will still be able to get reviewer feedback and push new patch sets, even with a -2. The list of core reviewers <https://review.openstack.org/#/admin/groups/352, members> for the specifications is small but mighty. (This is not necessarily the same list of core reviewers for code patches.)

Changes to existing specs

For approved but not-completed specs:

• cosmetic cleanup, fixing errors, and changing the definition of a feature can be done to the spec.

For approved and completed specs:

• changing a previously approved and completed spec should only be done for cosmetic cleanup or fixing errors.
• changing the definition of the feature should be done in a new spec.

Please see the Ironic specs process wiki page <https://wiki.openstack.org/ wiki/Ironic/Specs_Process> for further reference.