Bug 1417761

Summary: Huawei cinder driver is always enforcing QOS on backend side
Product: Red Hat OpenStack Reporter: nalmond
Component: openstack-cinderAssignee: Eric Harney <eharney>
Status: CLOSED CURRENTRELEASE QA Contact: Tzach Shefi <tshefi>
Severity: medium Docs Contact:
Priority: medium    
Version: 9.0 (Mitaka)CC: eharney, jniu, pgrist, scohen, srevivo, zhouweigang09
Target Milestone: ---Keywords: OtherQA, Triaged, ZStream
Target Release: 12.0 (Pike)   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of:
: 1561746 (view as bug list) Environment:
Last Closed: 2019-07-10 15:43:48 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On: 1561746    
Bug Blocks:    

Description nalmond 2017-01-30 21:57:49 UTC
Description of problem:
When using QOS for cinder volume, the huawei cinder driver attempts to enforce QOS on the backend even if consumer is set to front-end (nova). If the SAN is not able to do QOS, volume creation fails.

Version-Release number of selected component (if applicable):
openstack-cinder-8.1.1-2

How reproducible:
Every time a volume with QOS is created

Steps to Reproduce:
1. Use Huawei backend that doesn't support QOS
2. Set QOS consumer to front-end
3. Create volume

Actual results:
Volume creation fails

Expected results:
Volume creation succeeds

Additional info:
Modifying the driver by adding: https://github.com/huaweistorage/OpenStack_Driver/blob/master/Cinder/Mitaka/smartx.py#L48-L50 allows this to function correctly.

Comment 3 zhouweigang09 2017-07-21 07:51:03 UTC
Hi,

Huawei Cinder driver has fix this on Mitaka, Please check the path:
https://review.openstack.org/#/c/485489

BTW, for Openstack Newton and Ocata, I will provide the fixed path link soon.

Comment 4 zhouweigang09 2017-07-25 01:34:47 UTC
For openstack Ocata to fix Qos, Please find the path below:
https://review.openstack.org/#/c/486597/

Comment 6 Eric Harney 2017-07-26 13:07:08 UTC
(In reply to zhouweigang09 from comment #3)
> Huawei Cinder driver has fix this on Mitaka, Please check the path:
> https://review.openstack.org/#/c/485489
> 
> BTW, for Openstack Newton and Ocata, I will provide the fixed path link soon.

This change needs to be backported to upstream driverfixes/newton as well before we can land it in OSP9.  Otherwise there is a regression for users updating from OSP9 to OSP10 if we fix this only in OSP9.

Comment 7 zhouweigang09 2017-07-28 01:31:18 UTC
(In reply to Eric Harney from comment #6)
> (In reply to zhouweigang09 from comment #3)
> > Huawei Cinder driver has fix this on Mitaka, Please check the path:
> > https://review.openstack.org/#/c/485489
> > 
> > BTW, for Openstack Newton and Ocata, I will provide the fixed path link soon.
> 
> This change needs to be backported to upstream driverfixes/newton as well
> before we can land it in OSP9.  Otherwise there is a regression for users
> updating from OSP9 to OSP10 if we fix this only in OSP9.


Hello Eric,

This fixed is backported to upstream driverfixes/newton in below path:
https://review.openstack.org/#/c/487680

Please go ahead.

Comment 10 Tzach Shefi 2018-04-04 06:07:51 UTC
Set OtherQA, no such backend to verify this on.