Bug 2004214 - NetApp ONTAP cannot resize a volume LUN with sub-clone
Summary: NetApp ONTAP cannot resize a volume LUN with sub-clone
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat OpenStack
Classification: Red Hat
Component: openstack-cinder
Version: 13.0 (Queens)
Hardware: Unspecified
OS: Linux
medium
medium
Target Milestone: z6
: 16.2 (Train on RHEL 8.4)
Assignee: Cinder Bugs List
QA Contact: Luigi Toscano
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2021-09-14 17:50 UTC by Eduardo Santos
Modified: 2023-11-08 19:18 UTC (History)
11 users (show)

Fixed In Version: openstack-cinder-15.6.1-2.20230906144854.299553a.el8ost
Doc Type: Bug Fix
Doc Text:
This update fixes a NetApp Block Storage (cinder) volume driver issue. Previously, a volume extend operation could fail when the extended size was greater than the maximum LUN geometry on the back end due to a malformed request from the driver. The driver now includes the correct information in the request.
Clone Of:
Environment:
Last Closed: 2023-11-08 19:18:30 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Launchpad 1924643 0 None None None 2021-09-17 04:30:11 UTC
OpenStack gerrit 798208 0 None NEW NetApp ONTAP: Fix sub-clone zapi call 2021-11-11 00:41:49 UTC
Red Hat Issue Tracker OSP-9596 0 None None None 2021-11-15 12:50:37 UTC
Red Hat Product Errata RHBA-2023:6307 0 None None None 2023-11-08 19:18:56 UTC

Description Eduardo Santos 2021-09-14 17:50:30 UTC
Hi,

When trying to create a clone using Cinder with Netapp driver on OpenStack we encountered the following error:

Cannot set space reservation for block ranges. :: ONTAPI :: Error

This happens when trying to create a clone with snapshot of size larger than 2 TB, with the block ranges are being specified in the clone creation request.

Whereas, when trying to create a clone with snapshot of size smaller than 2 TB, the cloning request is coming to the storage without the block range specified and it completes successfully.

A bug report was already filed on Cinder's Launchpad [1] and a change has already been proposed to address the issue [2].

Rocky or older (including Queens) OpenStack releases are affected.

Steps to reproduce:
1. Deploy a queens release
2. Configure ONTAP 9.3
3. Create a volume smaller than 64G (e.g. openstack volume create --size 4 test)
4. Try to extend the volume to larger than 64G (e.g. openstack volume set test --size 65)

Actual result:
Volume is in error_extending state, and c-vol logs an error: http://paste.openstack.org/show/804536/

Expected result:
Volume is extended to the specified size (e.g. volume 'test' would be 65G)

[1] https://bugs.launchpad.net/cinder/+bug/1924643
[2] https://review.opendev.org/c/openstack/cinder/+/798208

Comment 1 Luigi Toscano 2021-09-16 17:36:34 UTC
(In reply to Eduardo Santos from comment #0)
> Hi,
> 
> When trying to create a clone using Cinder with Netapp driver on OpenStack
> we encountered the following error:
> 
> Cannot set space reservation for block ranges. :: ONTAPI :: Error
> 
> This happens when trying to create a clone with snapshot of size larger than
> 2 TB, with the block ranges are being specified in the clone creation
> request.
> 
> Whereas, when trying to create a clone with snapshot of size smaller than 2
> TB, the cloning request is coming to the storage without the block range
> specified and it completes successfully.
> 
> A bug report was already filed on Cinder's Launchpad [1] and a change has
> already been proposed to address the issue [2].
> 
> Rocky or older (including Queens) OpenStack releases are affected.

Please note that Red Hat OpenStack Platform 13  (based on Queens) is now in Extended Lifecycle Support (ELS) state, where only a limited amount of fix is applied, please check:
 https://access.redhat.com/support/policy/updates/openstack/platform

Is this bug (while waiting for the fix to be merged in opendev.org) still relevant for OSP13, or can it be targeted to the more recent 16.x releases?

Comment 7 Nahim Souza 2021-10-06 14:52:47 UTC
(In reply to Luigi Toscano from comment #1)
> (In reply to Eduardo Santos from comment #0)
> > Hi,
> > 
> > When trying to create a clone using Cinder with Netapp driver on OpenStack
> > we encountered the following error:
> > 
> > Cannot set space reservation for block ranges. :: ONTAPI :: Error
> > 
> > This happens when trying to create a clone with snapshot of size larger than
> > 2 TB, with the block ranges are being specified in the clone creation
> > request.
> > 
> > Whereas, when trying to create a clone with snapshot of size smaller than 2
> > TB, the cloning request is coming to the storage without the block range
> > specified and it completes successfully.
> > 
> > A bug report was already filed on Cinder's Launchpad [1] and a change has
> > already been proposed to address the issue [2].
> > 
> > Rocky or older (including Queens) OpenStack releases are affected.
> 
> Please note that Red Hat OpenStack Platform 13  (based on Queens) is now in
> Extended Lifecycle Support (ELS) state, where only a limited amount of fix
> is applied, please check:
>  https://access.redhat.com/support/policy/updates/openstack/platform
> 
> Is this bug (while waiting for the fix to be merged in opendev.org) still
> relevant for OSP13, or can it be targeted to the more recent 16.x releases?

Hi Luigi, we understand that it's relevant for OSP13 because though some of our customers already use OSP16, most of them still use OSP13 and they can have problems in some basic functions like cloning and extending volumes.

Comment 9 Alan Bishop 2021-11-01 12:58:07 UTC
Posting public comment (most of the private comments in this BZ do not need to be private). Also responding for Tosky, who is on PTO.

Netapp has posted a patch upstream, and it has a couple of +1 votes. It it Netapp's responsibility for advocating for additional reviews, especially from core reviewers. This can be done in the #openstack-cinder irc channel, or in the weekly cinder meeting (also on irc). A good way of prompting for reviewers is to mention the patch is in a good state (all comments addressed, and CI passed), and customers are waiting for the fix.

However, looking at the upstream gerrit results indicate Netapp's own CI isn't passing. Netapp needs to monitor the situation, and take the steps to get it working. Core reviewers will not approve driver patches unless the driver's own CI tests are passing.

Once the patch has merged on upstream master, upstream backports need to be posted sequentially for all relevant upstream releases (xena, wallaby, victoria, ussuri, train, etc.).

Comment 10 fabiooliveira 2021-11-22 14:01:38 UTC
Hello, we got the patch merged  into master and now we are working on the backports

https://review.opendev.org/c/openstack/cinder/+/798208

Comment 11 fabiooliveira 2022-02-15 14:22:07 UTC
Hello,

Backports to Xena (https://review.opendev.org/c/openstack/cinder/+/821890) and Wallaby (https://review.opendev.org/c/openstack/cinder/+/821892) already got merged.

We will provide a manual QA as a requirement to merge this bugfix on the remaining stable branches until Train.

Comment 18 Brian Rosmaita 2023-10-18 17:02:29 UTC
Added suggested documentation.

Comment 27 errata-xmlrpc 2023-11-08 19:18:30 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Red Hat OpenStack Platform 16.2.6 (Train) bug fix and enhancement advisory), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2023:6307


Note You need to log in before you can comment on or make changes to this bug.