Cinder manage now requires a pool name, which administrators have no way to discover
Bug #1364279 reported by
Navneet
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Cinder |
Fix Released
|
Low
|
Huang Zhiteng |
Bug Description
The new cinder pools submission breaks any cinder operation that involves <host> as input parameter. I tested cinder manage/unmanage feature taking the latest code. cinder manage breaks if host not given in the form host@backend#pool. For eg the following breaks:
cinder manage --source-name vol1 openstack1@iscsi
but this succeeds
cinder manage --source-name vol1 openstack1@
The knowledge of pool is only limited to the backend and hence any operation that involves <host> as input should not be burdened to also include pool name with it. This should be looked into at high priority.
tags: | removed: netapp |
Changed in cinder: | |
status: | New → Confirmed |
Changed in cinder: | |
status: | Confirmed → Invalid |
summary: |
- Cinder pools breaks operations involving host + Cinder manage now requires a pool name, which administrators have no way + to discover |
Changed in cinder: | |
assignee: | nobody → Huang Zhiteng (zhiteng-huang) |
status: | Confirmed → In Progress |
Changed in cinder: | |
importance: | Undecided → Low |
Changed in cinder: | |
milestone: | none → juno-rc1 |
status: | Fix Committed → Fix Released |
Changed in cinder: | |
milestone: | juno-rc1 → 2014.2 |
To post a comment you must log in.
I'm able to reproduce this behavior with DevStack-LVM setting.
Only something like
$ cinder manage --source-name volume- a618c57e- acde-40a3- bda9-cc12cbef07 58 ra@lvmdriver- 1#lvmdriver- 1
succeeds. If the part starting from '#' is not specified, the volume appears in the 'cinder list', but is in an 'error' state. And in the scheduler log, it says:
2014-09-04 20:58:41.075 ERROR cinder. scheduler. manager [req-f55acb5b- 58db-4805- 9fd2-0e22ef5d2b 6e ddd702b542b44d3 0a871abe64eabd7 81 a9c3b3cc2f554ef 88115b2262f2e39 3c] Failed to schedule_ manage_ existing: No valid host was found. Cannot place volume ecc02267- c3ba-4d42- 8323-1ccd14c10b b2 on ra@lvmdriver-1
Another thing I found (maybe it constitutes another bug) is when the host is not in proper format (<hostname> @<backend- name>), or if there is a typo in specifying the host, the response error message is 'Service not found'. The word 'service' is slightly confusing IMO. 'Backend not found', or 'host not found' would make more sense...