Volume backend with several pools may return wrong backend state
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Cinder |
In Progress
|
Undecided
|
Unassigned |
Bug Description
If volume backend is consists of two pools,one of the pool's backend state is up and another is down, the backend state would be down.
For example:
Two pools:
{u'name': u'cinder@
{u'name': u'cinder@
But backend_
[root@2c510 tecs(keystone_
+------
| Binary | Host | Zone | Status | State | Updated_at | Disabled Reason | Backend State |
+------
| cinder-scheduler | cinder | nova | enabled | up | 2018-03-
| cinder-volume | cinder@ZTE_FCSAN-2 | nova | enabled | up | 2018-03-
+------
[root@2c510 tecs(keystone_
+------
| Property | Value |
+------
| QoS_support | True |
| allocated_
| backend_state | up |
| contiguous_
| driver_version | 6.0.0 |
| free_capacity_gb | 2609 |
| is_pool | True |
| location_info | 000000000000 |
| max_over_
| multiattach | True |
| name | cinder@
| pool_name | pool1 |
| provisioned_
| reserved_percentage | 0 |
| storage_protocol | iSCSI |
| thick_provision
| thin_provisioni
| timestamp | 2018-03-
| total_capacity_gb | 2775 |
| vendor_name | ZTE |
| volume_backend_name | FCSAN_2 |
+------
+------
| Property | Value |
+------
| QoS_support | True |
| allocated_
| backend_state | down |
| contiguous_
| driver_version | 6.0.0 |
| free_capacity_gb | 0 |
| is_pool | True |
| location_info | 000000000000 |
| max_over_
| multiattach | True |
| name | cinder@
| pool_name | pool111 |
| provisioned_
| reserved_percentage | 0 |
| storage_protocol | iSCSI |
| thick_provision
| thin_provisioni
| timestamp | 2018-03-
| total_capacity_gb | 0 |
| vendor_name | ZTE |
| volume_backend_name | FCSAN_2 |
+------
Changed in cinder: | |
assignee: | nobody → yuyafei (yu-yafei) |
description: | updated |
Changed in cinder: | |
assignee: | Yafei Yu (yu-yafei) → nobody |
Fix proposed to branch: master /review. openstack. org/553651
Review: https:/