[Openstack] Cinder volume teardown bug (maybe in TGTD?)
David Hill
david.hill at ubisoft.com
Sat Nov 2 01:15:47 UTC 2013
Hi guys,
Some times when we delete volumes, I've noticed that tgtd seems to keep a lock on the LVM.
I've been able to force delete the target from tgtd but I still cannot delete the logical volume. Seems
like a bug in tgtd. Did you ever experience this behavior?
Thank you,
Dave
We can see the session was disconnected from the compute (well, maybe not totally but we don't
see it any longer in iscsiadm sessions...
[root at compute05 ~]# iscsiadm -m session
tcp: [21] 10.130.0.11:3260,1 iqn.2010-10.org.openstack:volume-892ef2b5-57b8-4c2e-9e54-050da70418b1
But in tgtd, it's shown as still connected:
[root at cinder01 ~]# tgt-admin -s | less
Target 33: iqn.2010-10.org.openstack:volume-4bd794a3-611c-444f-9ca7-891f0e471b41
System information:
Driver: iscsi
State: ready
I_T nexus information:
I_T nexus: 63
Initiator: iqn.1994-05.com.redhat:cce2bb2d6ea5
Connection: 0
IP Address: 10.1.2.3
LUN information:
LUN: 0
Type: controller
SCSI ID: IET 00210000
SCSI SN: beaf330
Size: 0 MB, Block size: 1
Online: Yes
Removable media: No
Prevent removal: No
Readonly: No
Backing store type: null
Backing store path: None
Backing store flags:
LUN: 1
Type: disk
SCSI ID: IET 00210001
SCSI SN: beaf331
Size: 107374 MB, Block size: 512
Online: Yes
Removable media: No
Prevent removal: No
Readonly: No
Backing store type: rdwr
Backing store path: /dev/cinder-volumes/volume-4bd794a3-611c-444f-9ca7-891f0e471b41
Backing store flags:
Account information:
ACL information:
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20131101/56c29632/attachment.html>
More information about the Openstack
mailing list