Ironic

_do_takeover not taking node provision state into account

Bug #1421370 reported by Chris Krelle
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Ironic
Fix Released
High
Chris Krelle
Ironic 2015.1.0 "kilo"

Bug Description

When a conductor takes over for another node the conductor doing the takeover is not taking the nodes provision state in to account. this results in a incorrect pxe config where the pxe boot mode is set to deploy for active nodes. _do_takeover should check the provision state of the node it is taking over and set the default pxe boot to deploy or boot based on the nodes last known state.

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to ironic (master)

Fix proposed to branch: master
Review: https://review.openstack.org/155460

Changed in ironic:
assignee: nobody → Devananda van der Veen (devananda)
status: New → In Progress
OpenStack Infra (hudson-openstack)
Changed in ironic:
assignee: Devananda van der Veen (devananda) → David Shrewsbury (dshrews)
Dmitry Tantsur (divius)
Changed in ironic:
importance: Undecided → High
OpenStack Infra (hudson-openstack)
Changed in ironic:
assignee: David Shrewsbury (dshrews) → moorryan (moorryan)
OpenStack Infra (hudson-openstack)
Changed in ironic:
assignee: moorryan (moorryan) → Chris Krelle (nobodycam)
OpenStack Infra (hudson-openstack)
Changed in ironic:
assignee: Chris Krelle (nobodycam) → Alexis Lee (alexisl)
OpenStack Infra (hudson-openstack)
Changed in ironic:
assignee: Alexis Lee (alexisl) → Chris Krelle (nobodycam)
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to ironic (master)

Reviewed: https://review.openstack.org/155460
Committed: https://git.openstack.org/cgit/openstack/ironic/commit/?id=bf75b8709b93cc41779f538249347a0dd07b6b14
Submitter: Jenkins
Branch: master

commit bf75b8709b93cc41779f538249347a0dd07b6b14
Author: Devananda van der Veen <email address hidden>
Date: Thu Feb 12 11:38:04 2015 -0800

    Correctly rebuild the PXE file during takeover of ACTIVE nodes

    During a conductor take-over, eg. from another failed conductor, it
    rebuilds the PXE configuration of nodes that are mapped to it. This
    calls node.driver.deploy.prepare(). However, in the PXE driver,
    prepare() was not properly handling the case where it was called on a
    node with an ACTIVE deploy state.

    In such a situation, prepare() needs to switch from the "deploy" config
    to the "boot" config. Furthermore, the PXE boot file must specify the
    node's root disk UUID. This was previously not available outside of the
    _continue_deploy() method, and so is now saved on
    node.driver_internal_info when the initial deploy is completed.

    Co-Authored-By: David Shrewsbury <email address hidden>
    Co-Authored-By: Chris Krelle <email address hidden>
    Change-Id: I6adc65b756e22c5c78d30231187fe2e8ccf2594d
    Closes-bug: #1421370

Changed in ironic:
status: In Progress → Fix Committed
Thierry Carrez (ttx)
Changed in ironic:
milestone: none → kilo-3
status: Fix Committed → Fix Released
Thierry Carrez (ttx)
Changed in ironic:
milestone: kilo-3 → 2015.1.0
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.