A restore from backup ought to keep the sparse volumes

Bug #2007615 reported by Pete Zaitcev
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Cinder
Fix Released
Medium
Pete Zaitcev

Bug Description

At some installations, a significant fraction of volumes is sparse. If such a cluster has to be restored in a disaster recovery, it may not be possible to restore into the exact same equipment that was in the datacenter originally. As Cinder restores, it unrolls backup images with no regard to sparseness, and thus uses more space than was used at the time of backup.

Note that the backup saves the volumes correctly, preserving sparseness as compressed zeroes. The loss of sparseness occurs during the restore process.

Pete Zaitcev (zaitcev)
Changed in cinder:
assignee: nobody → Pete Zaitcev (zaitcev)
Revision history for this message
Pete Zaitcev (zaitcev) wrote :

Red Hat had a bug about literally the same issue that was open back in 2020:
https://bugzilla.redhat.com/show_bug.cgi?id=1855709

Changed in cinder:
status: New → In Progress
Changed in cinder:
importance: Undecided → Medium
tags: added: backup-service
Revision history for this message
Rajat Dhasmana (whoami-rajat) wrote :
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to cinder (stable/2023.1)

Fix proposed to branch: stable/2023.1
Review: https://review.opendev.org/c/openstack/cinder/+/877799

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

Reviewed: https://review.opendev.org/c/openstack/cinder/+/852654
Committed: https://opendev.org/openstack/cinder/commit/b75c29c7d8e0e6ac212b59f9ad8d140874e55251
Submitter: "Zuul (22348)"
Branch: master

commit b75c29c7d8e0e6ac212b59f9ad8d140874e55251
Author: Pete Zaitcev <email address hidden>
Date: Wed Jul 20 12:15:21 2022 -0500

    Restore into sparse volumes

    We take the approach to detect zero data when restoring and not
    writing it out to the volume. This works exactly as expected
    when restoring into Linux LVM volumes and into Ceph RBD volumes,
    provided they are freshly created.

    Closes-bug: #2007615
    Change-Id: I54b81a568a01af44e3f74bcac55e823cdae9bfbf

Changed in cinder:
status: In Progress → Fix Released
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to cinder (stable/2023.1)

Reviewed: https://review.opendev.org/c/openstack/cinder/+/877799
Committed: https://opendev.org/openstack/cinder/commit/c56b79caa3958dec47f6aa1815abc7638a42952a
Submitter: "Zuul (22348)"
Branch: stable/2023.1

commit c56b79caa3958dec47f6aa1815abc7638a42952a
Author: Pete Zaitcev <email address hidden>
Date: Wed Jul 20 12:15:21 2022 -0500

    Restore into sparse volumes

    We take the approach to detect zero data when restoring and not
    writing it out to the volume. This works exactly as expected
    when restoring into Linux LVM volumes and into Ceph RBD volumes,
    provided they are freshly created.

    Closes-bug: #2007615
    Change-Id: I54b81a568a01af44e3f74bcac55e823cdae9bfbf
    (cherry picked from commit 477b1342747ddeb3ab9a586b33c06f1e51f37df8)

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix included in openstack/cinder 22.0.0.0rc2

This issue was fixed in the openstack/cinder 22.0.0.0rc2 release candidate.

Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix included in openstack/cinder 23.0.0.0rc1

This issue was fixed in the openstack/cinder 23.0.0.0rc1 release candidate.

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.