Bug 1826365

Summary: LSM for ISCSI disk size smaller than 1GB created with API fails
Product: [oVirt] ovirt-engine Reporter: Ilan Zuckerman <izuckerm>
Component: BLL.StorageAssignee: Benny Zlotnik <bzlotnik>
Status: CLOSED CURRENTRELEASE QA Contact: Evelina Shames <eshames>
Severity: high Docs Contact:
Priority: medium    
Version: 4.4.0CC: aefrat, aoconnor, bugs, dfodor, lsvaty, michal.skrivanek, mtessun, sfishbai
Target Milestone: ovirt-4.4.3Flags: pm-rhel: ovirt-4.4+
aoconnor: blocker-
mtessun: planning_ack+
sfishbai: testing_ack+
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: No Doc Update
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2020-11-11 06:42:28 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: Storage RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
engine log
none
vdsm log
none
4.3 env logs none

Description Ilan Zuckerman 2020-04-21 14:18:23 UTC
Created attachment 1680575 [details]
engine log

Description of problem:

When making a live disk migration between two ISCSI domains of disk size smaller that 1GB (created with API), The vdsm fails to migrate with:

2020-04-21 16:59:33,672+0300 ERROR (tasks/2) [storage.Volume] The requested initial 976130048 is bigger than the max size 134217728 (blockVolume:339)

Causing more errors and exceptions such as:

vdsm.storage.exception.InvalidParameterException: Invalid parameter: 'initial_size=976130048'

Engine is failing on:

2020-04-21 16:59:35,883+03 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedScheduledExecutorService-engineScheduledThreadPool-Thread-93) [] EVENT_ID: VDS_BROKER_COMMAND_FAILURE(10,802), VDSM host_mixed_1 command HSMGetAllTasksStatusesVDS failed: value=Error creating a new volume: ("Volume creation 97ea7f1f-4497-4d42-94bd-f0e409f15736 failed: Invalid parameter: 'initial_size=976130048'",) abortedcode=205


Tested exactly the same scenario on earlier 4.3 build, and it succeeded:
ovirt-engine-4.3.9.4-11.el7.noarch
vdsm-4.30.44-1.el7ev.x86_64


Version-Release number of selected component (if applicable):
vdsm-4.40.13-1.el8ev.x86_64
ovirt-engine-4.4.0-0.33.master.el8ev.noarch

How reproducible:
100%

Steps to Reproduce:

1. Create blank vm
2. Create Qcow disk with size smaller that 1GB + attach it to the vm as os disk on ISCSI

POST {{engine}}/disks/

BODY:
<disk>
  <storage_domains>
    <storage_domain id="{{underlying_sd}}"/>
  </storage_domains>
  <name>{{tc_name}}_qcow</name>
  <provisioned_size>1048576</provisioned_size>
  <format>cow</format>
  <sparse>true</sparse>
</disk>

3. Start the vm -> wait till it starts
4. Migrate the disk to another ISCSI domain


Actual results:
Migration fails

Expected results:
Should succeed as it does on 4.3


Additional info:
Attaching engine logs + vdsm log (which is also SPM)

Comment 1 Ilan Zuckerman 2020-04-21 14:18:49 UTC
Created attachment 1680577 [details]
vdsm log

Comment 2 RHEL Program Management 2020-04-22 07:56:44 UTC
This bug report has Keywords: Regression or TestBlocker.
Since no regressions or test blockers are allowed between releases, it is also being identified as a blocker for this release. Please resolve ASAP.

Comment 3 Benny Zlotnik 2020-04-22 10:05:34 UTC
any chance you have the logs from 4.3 too?

Comment 4 Ilan Zuckerman 2020-04-22 12:25:15 UTC
(In reply to Benny Zlotnik from comment #3)
> any chance you have the logs from 4.3 too?

yes. attaching.

Comment 5 Ilan Zuckerman 2020-04-22 12:26:09 UTC
Created attachment 1680847 [details]
4.3 env logs

Comment 6 Benny Zlotnik 2020-04-22 14:08:12 UTC
I looked into this, it's not really a regression, it doesn't work in 4.3 as well. I checked this with cold move (there has to be a snapshot created though) and it fails in both versions. The reason LSM only fails in 4.4 is that in 4.4 LSM uses the same copy/move mechanism as cold move (engine controlled vs. vdsm controlled)

Comment 7 Evelina Shames 2020-10-13 11:56:06 UTC
Verified on rhv-4.4.3-8.

Comment 8 Sandro Bonazzola 2020-11-11 06:42:28 UTC
This bugzilla is included in oVirt 4.4.3 release, published on November 10th 2020.

Since the problem described in this bug report should be resolved in oVirt 4.4.3 release, it has been closed with a resolution of CURRENT RELEASE.

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