Bug 1252575 - NetApp NFS drivers ignore nfs_oversub_ratio when reporting pool capacity
Summary: NetApp NFS drivers ignore nfs_oversub_ratio when reporting pool capacity
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat OpenStack
Classification: Red Hat
Component: openstack-cinder
Version: 6.0 (Juno)
Hardware: Unspecified
OS: Unspecified
urgent
urgent
Target Milestone: async
: 6.0 (Juno)
Assignee: Eric Harney
QA Contact: nlevinki
URL:
Whiteboard:
: 1266552 (view as bug list)
Depends On: 1251115
Blocks:
TreeView+ depends on / blocked
 
Reported: 2015-08-11 18:36 UTC by Eric Harney
Modified: 2023-09-14 03:03 UTC (History)
14 users (show)

Fixed In Version: openstack-cinder-2014.2.4-2.el7ost
Doc Type: Rebase: Bug Fixes Only
Doc Text:
Previously, NetApp NFS drivers, both cDOT and 7mode, ignored nfs_oversub_ratio when reporting pool capacity. These drivers did not make use of the nfs_oversub_ratio configuration parameter to enable OpenStack administrators to report anything other than the most conservative capacity and usage information up to the scheduler. Since these drivers inherit from the generic NFS driver, where nfs_oversub_ratio can be configured, they allowed administrators to configure over-subscription, but did not actually implement it. This fix now uses nfs_oversub_ratio when reporting pool capacity, and: -modifies the NetApp drivers to use the nfs_oversub_ratio as documented -uses direct API with filers to gather capacity information rather than stat and du commands -brings our reporting of reserved percentage in line with the way the scheduler actually makes use of this statistic -simplifies and cleans up the way we gather and report pool statistics
Clone Of: 1251115
Environment:
Last Closed: 2016-03-03 19:31:17 UTC
Target Upstream Version:
Embargoed:

Attachments (Terms of Use)
Modified script by customer (69.59 KB, text/plain)
2016-01-23 05:13 UTC, Shinobu KINJO 		no flags Details
View All


Links
System ID Private Priority Status Summary Last Updated
OpenStack gerrit 172172 0 None MERGED Only use operational LIFs for iscsi target details 2020-06-23 05:29:13 UTC
OpenStack gerrit 185724 0 None MERGED Use nfs_oversub_ratio when reporting pool capacity 2020-06-23 05:29:13 UTC
OpenStack gerrit 225185 0 None MERGED Fix NetApp clone from glance failure 2020-06-23 05:29:13 UTC
Red Hat Product Errata RHBA-2016:0353 0 normal SHIPPED_LIVE openstack-cinder bug fix advisory 2016-03-04 00:30:46 UTC

Comment 4 Sergey Gotliv 2015-09-27 13:40:00 UTC 
According to BZ#1266552, we also have to backport [1] to fix the issue described here.

[1] https://review.openstack.org/#/c/172172/1
Comment 5 Sergey Gotliv 2015-09-27 13:44:17 UTC 
*** Bug 1266552 has been marked as a duplicate of this bug. ***
Comment 6 Lee Yarwood 2015-10-01 11:51:26 UTC 
In a further twist this also appears to introduce another issue where by netapp copy offloading no longer works :

~~~
2015-10-01 10:25:18.498 8329 INFO cinder.volume.drivers.netapp.nfs [req-35bbd494-5904-4e85-aac4-a9fa78e3c8d2 ce6b3728d50a48508e26431f6ae92a27 9e9127203fef452cb93863b211729ac6 - - -] Cloning image ca6b9576-8075-4741-a6bb-af985625b69a from cache
2015-10-01 10:25:19.057 8329 INFO cinder.volume.drivers.netapp.nfs [req-35bbd494-5904-4e85-aac4-a9fa78e3c8d2 ce6b3728d50a48508e26431f6ae92a27 9e9127203fef452cb93863b211729ac6 - - -] Image cloning unsuccessful for image ca6b9576-8075-4741-a6bb-af985625b69a. Message: need more than 2 values to unpack
~~~

Reported upstream [1], fixed in Liberty with another patch pending for stable/kilo [2], I'll take a look at getting this back into stable/juno for another test build now.

Lee

[1] https://bugs.launchpad.net/cinder/+bug/1490845
[2] https://review.openstack.org/#/c/225185/
Comment 7 Sergey Gotliv 2015-10-06 11:51:11 UTC 
Lee/Dave,

According to the previous comment the customer has more issues with the NetApp driver than we expected. Please, communicate that bug to the NetApp because we'll have to backport more patches.
Comment 12 Sergey Gotliv 2015-10-17 21:02:21 UTC 
NetApp folks backported additional patches to their github branch. IIUC, they did something very similar to what Lee proposed in the hotfix. NetApp still have to pass their official process in order to make these patches available. Let's see what happens first, NetApp release their patches or we get a feedback from the customer.

Meantime, I will ask them to review Lee's work.
Comment 17 Sergey Gotliv 2015-12-15 12:31:26 UTC 
This patch [1] has to be merged to close that bug.

[1] https://code.engineering.redhat.com/gerrit/61706
Comment 34 Shinobu KINJO 2016-01-23 05:13:52 UTC 
Created attachment 1117408 [details]
Modified script by customer
Comment 45 Shinobu KINJO 2016-02-24 05:59:06 UTC 
This bugzilla mentioned that 2014.2.4-2 is fixed version.

 Fixed In Version: openstack-cinder-2014.2.4-2.el7ost

But errata show us 2014.2.3-11 as fixed version.

 openstack-cinder-2014.2.3-11.el7ost.noarch.rpm

  1.Details: https://goo.gl/SZEC06
  2.Change Logs: https://goo.gl/Dk74cq

What I would like to be informed is that:

 1. Is there any reason about this difference?
 2. What is the difference between 2 versions?

FYI: customer confirmed that the 2014.2.4 fixed the issue and that version containing the necessary fixes.
Please update as soon as possible.

Rgds,
Comment 56 nlevinki 2016-02-29 14:10:12 UTC 
code is in
openstack-cinder-2014.2.4-2.el7ost.noarch
automation passed
https://rhos-jenkins.rhev-ci-vms.eng.rdu2.redhat.com/view/RHOS/view/RHOS6/job/rhos-jenkins-rhos-6.0-puddle-rhel-7.2-multi-node-packstack-neutron-ml2-vxlan-rabbitmq-tempest-git-all/12/
Comment 58 errata-xmlrpc 2016-03-03 19:31:17 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, 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://rhn.redhat.com/errata/RHBA-2016-0353.html
Comment 59 Red Hat Bugzilla 2023-09-14 03:03:34 UTC 
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days
Note You need to log in before you can comment on or make changes to this bug.