Note: This is a beta release of Red Hat Bugzilla 5.0. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Also email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback here.
Bug 1688507 - ceph-ansible should support setting device classes on OSDs
Summary: ceph-ansible should support setting device classes on OSDs
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Red Hat Ceph Storage
Classification: Red Hat
Component: Ceph-Ansible
Version: 3.1
Hardware: Unspecified
OS: Unspecified
unspecified
medium
Target Milestone: rc
: 4.0
Assignee: Dimitri Savineau
QA Contact: ceph-qe-bugs
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2019-03-13 21:17 UTC by Tupper Cole
Modified: 2019-03-18 21:19 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2019-03-18 18:29:11 UTC
Target Upstream Version:


Attachments (Terms of Use)

Description Tupper Cole 2019-03-13 21:17:40 UTC
Description of problem:
ceph-ansible should support the setting of device classes at deployment on individual OSDs. 



Version-Release number of selected component (if applicable):3.2

The code looks like it is in place, but it doesn't seem to work properly.



https://github.com/ceph/ceph-ansible/blob/f2dcb02d213e862c5a5498c2d12cd86b22676c84/library/ceph_volume.py

Comment 1 Dimitri Savineau 2019-03-18 15:07:07 UTC
Do you have an exemple or ansible log of what isn't working ?

https://github.com/ceph/ceph-ansible/blob/stable-3.2/docs/source/osds/scenarios.rst#simple-configuration

------------
osd_scenario: lvm
devices:
  - /dev/sda
  - /dev/sdb
  - /dev/sdc
  - /dev/sdd
crush_device_class: foo
------------

foo device class will be applied to all OSDs.

https://github.com/ceph/ceph-ansible/blob/stable-3.2/docs/source/osds/scenarios.rst#advanced-configuration

------------
osd_scenario: lvm
lvm_volumes:
  - data: /dev/sda
    crush_device_class: foo
  - data: /dev/sdb
    crush_device_class: foo
  - data: /dev/sdc
    crush_device_class: bar
  - data: /dev/sdd
    crush_device_class: bar
------------

Set device_class per OSDs

Comment 4 Tupper Cole 2019-03-18 21:19:13 UTC
So the customer has to create the volume groups and logical volumes before deploying? This makes a 900 OSD cluster *rather* difficult to deploy, no?


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