One of the "fun" aspects of accessing OCI images, is we have no way to realistically gain awareness of the underlying disk format in the OCI model, at least unless it is hinted at in the data model. Where we're unable to really figure that out is when a user supplies a specific digest URL. Ironic recognizes this and "right sizes" the process and data discovery and explicitly notes the disk format it believe to be 'unknown'. In order for IPA to be able to stream, and appropriately check this data format, IPA has be "okay" with 'unknown'. Everything else appears good to get to this point. This doesn't prohibit the image safety checking, just allows for the perception mismatch when the format is 'unknown' Change-Id: Ibe38245e906c659057a3c5ea7d8a0e474599ff5c
Ironic Python Agent
Team and repository tags
Overview
An agent for controlling and deploying Ironic controlled baremetal nodes.
The ironic-python-agent works with the agent driver in Ironic to provision the node. Starting with ironic-python-agent running on a ramdisk on the unprovisioned node, Ironic makes API calls to ironic-python-agent to provision the machine. This allows for greater control and flexibility of the entire deployment process.
The ironic-python-agent may also be used with the original Ironic pxe drivers as of the Kilo OpenStack release.
Building the IPA deployment ramdisk
For more information see the Image Builder section of the Ironic Python Agent developer guide.
Using IPA with devstack
This is covered in the Deploying Ironic with DevStack section of the Ironic dev-quickstart guide.
Project Resources
Project bugs are tracked on Launchpad:
Developer documentation can be found here:
Release notes for the project are available at:
https://docs.openstack.org/releasenotes/ironic-python-agent/
Source code repository for the project is located at:
- IRC channel:
-
#openstack-ironic on irc.oftc.net
To contribute, start here: Openstack: How to contribute.