Instance in Error state should allow reboot / rebuild

Bug #1183946 reported by Phil Day
10
This bug affects 2 people
Affects Status Importance Assigned to Milestone
OpenStack Compute (nova)
Fix Released
Medium
Phil Day

Bug Description

Currently reboot and rebuild are blocked as actions for an instance with a state of ERROR. but in most cases these are valid recovery operations.

Suggest the following logic:

Reboot is allowed from Error providing launched_at is not Null (i.e the instance has booted at once)

Rebuild is allowed form Error proving the task state is not scheduling, block_device_mapping, or networking (or it should be made to come with networking not having completed)

Tags: api compute
tags: added: api
Changed in nova:
status: New → Confirmed
importance: Undecided → Medium
tags: added: compute
Phil Day (philip-day)
Changed in nova:
assignee: nobody → Phil Day (philip-day)
Phil Day (philip-day)
Changed in nova:
status: Confirmed → In Progress
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to nova (master)

Fix proposed to branch: master
Review: https://review.openstack.org/32145

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

Reviewed: https://review.openstack.org/32145
Committed: http://github.com/openstack/nova/commit/99c51e34230394cadf0b82e364ea10c38e193979
Submitter: Jenkins
Branch: master

commit 99c51e34230394cadf0b82e364ea10c38e193979
Author: Phil Day <email address hidden>
Date: Fri Jun 7 15:18:03 2013 +0100

    Allow reboot or rebuild from vm_state=Error

    In general most operations are only valid on an
    instance that has booted successfully at least once
    so this change extends the instance state check
    logic to include evidence that the instance has
    been successfully started at least once.

    This enables more operations to be allowed in
    against instances in an Error state. For example
    reboot and rebuild are useful recover options for
    an instance which has reached an error state, but
    not if the instance failed during its initial build.

    With this change the only actions allowed on an
    instance which has never booted successfully are
    delete and force_delete. Soft delete is not
    allowed, as the restore is in effect a start unless
    there is specific virt driver support.

    In addition the following actions are now allowed
    against instances in an Error state providing the
    instance has booted at least once: Reboot, Rebuild,
    and Rescue.

    Fixes bug: 1183946

    Change-Id: I63fd8d2a182df5cf12754892e8076933b3b1e79f

Changed in nova:
status: In Progress → Fix Committed
Thierry Carrez (ttx)
Changed in nova:
milestone: none → havana-2
status: Fix Committed → Fix Released
Thierry Carrez (ttx)
Changed in nova:
milestone: havana-2 → 2013.2
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

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