Bug 1308919

Description of problem:

- We have an instance named "feb3-selinux" which has a 5GB volume attached to it.
=======================================
[root@feb3-selinux data1]# lsblk
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
vda    253:0    0   20G  0 disk
└─vda1 253:1    0   20G  0 part /
vdb    253:16   0  4.8G  0 disk /data1
[root@feb3-selinux data1]# date
Tue Feb 16 05:35:13 EST 2016
[root@feb3-selinux data1]# df -kh
Filesystem      Size  Used Avail Use% Mounted on
/dev/vda1        20G  859M   20G   5% /
devtmpfs        1.9G     0  1.9G   0% /dev
tmpfs           1.9G     0  1.9G   0% /dev/shm
tmpfs           1.9G   17M  1.9G   1% /run
tmpfs           1.9G     0  1.9G   0% /sys/fs/cgroup
/dev/vdb        4.6G   19M  4.3G   1% /data1
====================================

- Next, we try to detach the volume from the Horizon dashboard. The volume is detached from the instance, but the state on the horizon dashboard keeps as detaching even after very long time.

- But, the volume is not attached to the instance any longer.

[root@feb3-selinux data1]# lsblk
NAME   MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
vda    253:0    0  20G  0 disk
└─vda1 253:1    0  20G  0 part /

[root@feb3-selinux data1]# date
Tue Feb 16 05:42:34 EST 2016
[root@feb3-selinux data1]# df -kh
Filesystem      Size  Used Avail Use% Mounted on
/dev/vda1        20G  859M   20G   5% /
devtmpfs        1.9G     0  1.9G   0% /dev
tmpfs           1.9G     0  1.9G   0% /dev/shm
tmpfs           1.9G   17M  1.9G   1% /run
tmpfs           1.9G     0  1.9G   0% /sys/fs/cgroup
/dev/vdb        4.6G   19M  4.3G   1% /data1
[root@feb3-selinux data1]# cd /data1
[root@feb3-selinux data1]#
[root@feb3-selinux data1]# ls -lrth
ls: reading directory .: Input/output error
total 0

- The volume is not shown in the multipath O/p on the compute node.

- When the volume is in detaching state, i manually go and update the status of the volume as available. But in the attached to column , it still shows "attached to feb3-selinux", even when it is actually not.

- After this, surprisingly when i do a hard reboot of the isntance, it actually attaches that volume back to the instance.
Find the below O/p for your reference.
============================
[gaurav.MCKD-4041] ➤ ssh root.xx.xx
X11 forwarding request failed on channel 0
Last login: Tue Feb 16 05:11:03 2016 from xx.xx.xx.xx
[root@feb3-selinux ~]#
[root@feb3-selinux ~]# lsblk
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
vda    253:0    0   20G  0 disk
└─vda1 253:1    0   20G  0 part /
vdb    253:16   0  4.8G  0 disk
[root@feb3-selinux ~]#

[root@comp3 ~]# multipath -ll | grep 360002ac0000000000000006400008044
360002ac0000000000000006400008044 dm-13 3PARdata,VV


Version-Release number of selected component (if applicable):
- RHOS6(non-ha)
- HP 3PAR backend for cinder(cinder.volume.drivers.san.hp.hp_3par_fc.HP3PARFCDriver)

How reproducible:
100%


Steps to Reproduce:
1.
2.
3.

Actual results:
unable to detach volume from instance.

Expected results:
volume should be detached from the instance.

Additional info: