Cinder resize procedure should not require lock/unlock

Bug #1803772 reported by Maria Yousaf
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
StarlingX
Fix Released
Medium
Wei Zhou

Bug Description

Brief Description
-----------------
The cinder resize procedure does not match what is documented. The current procedure requires a lock/unlock for some labs/disk types. The lock/unlock should not be required.

Severity
--------
Major

Steps to Reproduce
------------------
Currently to resize cinder successfully, the user must do the following:

1. On standby controller, lock controller, enlarge cinder partition and unlock
2. Swact
3. On new standby controller, lock controller, enlarge cinder partition and unlock
4. Swact
5. Now both cinder-volume sizes should be increased

In the documentation, cinder resize is described as an in-service procedure. No lock/unlock requirement is listed.

Expected Behavior
------------------
No lock/unlock required

Actual Behavior
----------------
Lock/unlock required with certain disks

Reproducibility
---------------
Reproducible

System Configuration
--------------------
Any LVM system with slower disks

Branch/Pull Time/Commit
-----------------------
N/A

Timestamp/Logs
--------------
N/A

Maria Yousaf (myousaf)
description: updated
Ghada Khalil (gkhalil)
Changed in starlingx:
importance: Undecided → Medium
status: New → Triaged
assignee: nobody → Wei Zhou (wzhou007)
tags: added: stx.config
tags: added: stx.2019.03
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to stx-config (master)

Fix proposed to branch: master
Review: https://review.openstack.org/619051

Changed in starlingx:
status: Triaged → In Progress
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to stx-config (master)

Reviewed: https://review.openstack.org/619051
Committed: https://git.openstack.org/cgit/openstack/stx-config/commit/?id=e052d9c86073afb669d6388521b3961e945ea555
Submitter: Zuul
Branch: master

commit e052d9c86073afb669d6388521b3961e945ea555
Author: Wei Zhou <email address hidden>
Date: Tue Nov 20 12:46:08 2018 -0500

    Cinder-volumes Volume Group Resize Procedure Failed

    When the disk partition for the cinder-volumes VG is increased
    on the newly standby controller after controller swact, drbd sync
    starts before cinder-volumes VG is extended. When drbd sync starts,
    the drbd is in "SyncSource" state. But the code waits for drbd to
    be in "Connected" state. If that doesn't happen within 5 seconds,
    it will fail and won't continue to extend cinder-volumes VG.

    The solution is to allow cinder-volumes VG extension while drbd is
    in either "Connected" or "SyncSource" state.

    Change-Id: Ib164ce5fb433789d0806ac8d4faae9d675695b65
    Closes-bug: 1803772
    Signed-off-by: Wei Zhou <email address hidden>

Changed in starlingx:
status: In Progress → Fix Released
Ken Young (kenyis)
tags: added: stx.2019.05
removed: stx.2019.03
Ken Young (kenyis)
tags: added: stx.2.0
removed: stx.2019.05
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.