Description of problem:
Unable to make use of DRP(data reduction pool) feature, when creating volumes using this pool

Version-Release number of selected component (if applicable):
RHOSP 13
IBM Storwize 8.1.3.6

How reproducible:
Always

Steps to Reproduce:
cinder:

[premium]
image_volume_cache_enabled=True
image_volume_cache_max_count=50
image_volume_cache_max_size_gb=1000
san_ip=172.16.30.50
san_login=admin
san_private_key=/etc/cinder/ibm.key
storwize_svc_flashcopy_rate=80
storwize_svc_volpool_name=DRP
storwize_svc_vol_warning = 0
volume_backend_name=premium
volume_driver=cinder.volume.drivers.ibm.storwize_svc.storwize_svc_fc.StorwizeSVCFCDriver
backend_host=hostgroup

Actual results:

2020-10-19 12:12:16.973 38 ERROR oslo_messaging.rpc.server command: ['svctask', 'mkvdisk', '-name', u'"volume-310a0676-9d79-43eb-a148-b3da7ebe0a36"', '-mdiskgrp', u'"DRP"', '-iogrp', u'0', '-size', '20', '-unit', 'gb', '-rsize', '2%', '-autoexpand', '-warning', '0%', '-grainsize', '256', '-easytier', 'on']
2020-10-19 12:12:16.973 38 ERROR oslo_messaging.rpc.server stdout:
2020-10-19 12:12:16.973 38 ERROR oslo_messaging.rpc.server stderr: CMMVC9236E The pool specified is a data reduction pool. Volumes or volume copies which are thin provisioned and created from a data reduction pool can not use the -warning parameter.

Expected results:

Created Successfully

Additional info:

Customer has three backends from the same SAN, with two different pool configuration,

[ibm]
image_volume_cache_enabled=True
image_volume_cache_max_count=50
image_volume_cache_max_size_gb=1000
san_ip=172.16.30.50
san_login=admin
san_private_key=/etc/cinder/ibm.key
storwize_svc_flashcopy_rate=80
storwize_svc_volpool_name=Pool0
volume_backend_name=ibm
volume_driver=cinder.volume.drivers.ibm.storwize_svc.storwize_svc_fc.StorwizeSVCFCDriver
backend_host=hostgroup
rpc_response_timeout = 1800

[slow]
san_ip=172.16.30.50
san_login=admin
san_private_key=/etc/cinder/ibm.key
storwize_svc_volpool_name=Standard
storwize_svc_vol_compression = False
storwize_svc_vol_easytier = False
volume_backend_name=slow
volume_driver=cinder.volume.drivers.ibm.storwize_svc.storwize_svc_fc.StorwizeSVCFCDriver
backend_host=hostgroup

[premium]
image_volume_cache_enabled=True
image_volume_cache_max_count=50
image_volume_cache_max_size_gb=1000
san_ip=172.16.30.50
san_login=admin
san_private_key=/etc/cinder/ibm.key
storwize_svc_flashcopy_rate=80
storwize_svc_volpool_name=DRP
storwize_svc_vol_warning = 0
volume_backend_name=premium
volume_driver=cinder.volume.drivers.ibm.storwize_svc.storwize_svc_fc.StorwizeSVCFCDriver
backend_host=hostgroup

(overcloud) [stack@aev-20001 ~]$ openstack volume type show 7149b54e-a8aa-466e-aefd-b0fdcd061adb
--------------------------------------------------------+

--------------------------------------------------------+

--------------------------------------------------------+
(overcloud) [stack@aev-20001 ~]$ openstack volume type show e2997709-e151-412f-bac1-4f3172b819e2
--------------------------------------------------------+

--------------------------------------------------------+

--------------------------------------------------------+
(overcloud) [stack@aev-20001 ~]$ openstack volume type show 0d1e9f6f-f83c-4b1a-a979-ca39449cb21d
--------------------------------------------------------+

--------------------------------------------------------+

--------------------------------------------------------+

While creating Volume requested on SAN, Openstack passes command on the SAN, which mainly involve the arguments like,

Below are the logs showing the SSH command being run:
2020-10-19 13:11:47.721 39 DEBUG oslo_concurrency.processutils [req-4a2f7797-8e49-470b-8d96-bd9f641ccf71 60ecc8e7c65343b5a46a658eb7727a95 90810eab1c3e4adfb050aeb2260fc89b - - -] Running cmd (SSH): svctask mkvdisk -name "volume-cda2e330-cdd3-4c4a-821a-018c1e780f6e" -mdiskgrp "Standard" -iogrp 0 -size 20 -unit gb -rsize 2% -autoexpand -warning 0% -grainsize 256 -easytier off ssh_execute /usr/lib/python2.7/site-packages/oslo_concurrency/processutils.py:525

But while creating from the premium Volume type and SAN pool DRP, following error is observed,

2020-10-19 12:12:16.973 38 ERROR oslo_messaging.rpc.server command: ['svctask', 'mkvdisk', '-name', u'"volume-310a0676-9d79-43eb-a148-b3da7ebe0a36"', '-mdiskgrp', u'"DRP"', '-iogrp', u'0', '-size', '20', '-unit', 'gb', '-rsize', '2%', '-autoexpand', '-warning', '0%', '-grainsize', '256', '-easytier', 'on']
2020-10-19 12:12:16.973 38 ERROR oslo_messaging.rpc.server stdout:
2020-10-19 12:12:16.973 38 ERROR oslo_messaging.rpc.server stderr: CMMVC9236E The pool specified is a data reduction pool. Volumes or volume copies which are thin provisioned and created from a data reduction pool can not use the -warning parameter.

This is a standard IBM error,

Now we asked the CU if a manual operation can be passed while removing the (-warning) flag

Removing the -warning argument removed the error:
IBM_Storwize:Cluster_172.16.30.5:superuser>mkvdisk -name "volume-test-manual-cli" -mdiskgrp "DRP" -iogrp 0 -size 20 -unit gb -rsize 2% -autoexpand -compressed -easytier off
CMMVC9247E The pool specified is a data reduction pool. Thin or compressed volumes or volume copies in data reduction pools cannot have the easytier status set or changed.

However it seems we also cannot specify the "easytier" argument. So removing that also then succeeds in the volume being created:

IBM_Storwize:Cluster_172.16.30.5:superuser>mkvdisk -name "volume-test-manual-cli" -mdiskgrp "DRP" -iogrp 0 -size 20 -unit gb -rsize 2% -autoexpand -compressed
Virtual Disk, id [455], successfully created
IBM_Storwize:Cluster_172.16.30.5:superuser>

Is there any way that (-warning) and (-easytier) flag can be explicitly removed from the Openstack request?

I have found the upstream document which basically overwrites the parameters specified on the configuration,
https://docs.openstack.org/mitaka/config-reference/block-storage/drivers/ibm-storwize-svc-driver.html

Also asked the CU to change the volume type properties -

openstack volume type set --property capabilities:warning='<is> False' --property drivers:warning=0 --property volume_backend_name=premium premium

But the same condition is observed, when passing command to the SAN as a standard (-warning) and (-easytier) are always being passed.

Looking for a way to bypass these flags/condition.