OpenStack Compute (nova)

migration is not recheduled if the server originally booted with --availability-zone <az>:<host>

Bug #1845291 reported by Balazs Gibizer
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
OpenStack Compute (nova)
In Progress
Medium
Balazs Gibizer
Ocata
Confirmed
Low
Unassigned
Pike
Confirmed
Low
Unassigned
Queens
Confirmed
Medium
Unassigned
Rocky
Confirmed
Medium
Unassigned
Stein
Confirmed
Medium
Unassigned

Bug Description

Steps to reproduce
==================
1) boot a server with --availability-zone <az>:<host> This will force nova to boot the server on the given host
2) Try to migrate the server in a situation when the first destination host of the migration selected by the scheduler will fail (e.g. move_claim fails) but there are alternate hosts that could support the migration.

Expected result
===============

Migration is re-scheduled after the first failure and can succeed on an alternate destination.

Actual result
=============
Nova does not try to re-schedule the migration after the first failure. Server goes to ERROR state.

Balazs Gibizer (balazs-gibizer)
tags: added: migration reschdule
tags: added: reschedule
removed: reschdule
Changed in nova:
assignee: nobody → Balazs Gibizer (balazs-gibizer)
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Related fix proposed to nova (master)

Related fix proposed to branch: master
Review: https://review.opendev.org/684545

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

Fix proposed to branch: master
Review: https://review.opendev.org/684546

Revision history for this message
Matt Riedemann (mriedem) wrote :

This goes back to Newton:

https://github.com/openstack/nova/commit/76dfb4ba9fa0fed1350021591956c4e8143b1ce9

Changed in nova:
status: New → In Progress
importance: Undecided → Medium
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Related fix merged to nova (master)

Reviewed: https://review.opendev.org/684545
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=167e593ed54a670e485a4e8c1d62e08fbd065911
Submitter: Zuul
Branch: master

commit 167e593ed54a670e485a4e8c1d62e08fbd065911
Author: Balazs Gibizer <email address hidden>
Date: Tue Sep 24 17:45:23 2019 +0200

    Functional reproduction for bug 1845291

    Nova does not try to re-schedule a migration if the original boot was
    done with --availability-zone <az:host> or <az::node>.

    Related-Bug #1845291

    Change-Id: Id78627c5c08090de6220249a5f44d26bf32724af

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

Reviewed: https://review.opendev.org/684546
Committed: https://git.openstack.org/cgit/openstack/nova/commit/?id=64b99bd853ab3e6b64a04333017c39b21010b029
Submitter: Zuul
Branch: master

commit 64b99bd853ab3e6b64a04333017c39b21010b029
Author: Balazs Gibizer <email address hidden>
Date: Tue Sep 24 18:03:33 2019 +0200

    Reset forced_destination before migration at a proper time

    Resetting was in place but it was done after the retry filter is
    populated in the MigrationTask by the populate_retry call. This
    patch moves the reset code before the call to populate_retry as
    to allow retries.

    Change-Id: I8290e890a36cf5a8f409ab8a50e7c72f7ae15025
    Closed-Bug: #1845291

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.