Public bug reported:
Under heavy load, openiscsi will sometimes report connection errors when
using tgt:
Apr 8 22:02:44 node003-cont001 iscsid: Connection2:0 to [target:
iqn.2010-10.org.openstack:volume-b32d317e-4063-4e4f-963e-b0fdc29198f9,
portal: 172.21.1.1,3260] through [iface: default] is operational now
Apr 8 22:10:53 node003-cont001 iscsid: Kernel reported iSCSI connection 2:0
error (1021 - ISCSI_ERR_SCSI_EH_SESSION_RST: Session was dropped as a result of
SCSI error recovery) state (3)
Apr 8 22:10:56 node003-cont001 iscsid: connection2:0 is operational after
recovery (1 attempts)
Apr 8 22:18:47 node003-cont001 iscsid: Kernel reported iSCSI connection 2:0
error (1021 - ISCSI_ERR_SCSI_EH_SESSION_RST: Session was dropped as a result of
SCSI error recovery) state (3)
Apr 8 22:18:50 node003-cont001 iscsid: connection2:0 is operational after
recovery (1 attempts)
Even though openiscsi has disconnected:
Apr 8 22:34:04 node003-cont001 iscsid: Connection2:0 to [target:
iqn.2010-10.org.openstack:volume-b32d317e-4063-4e4f-963e-b0fdc29198f9,
portal: 172.21.1.1,3260] through [iface: default] is shutdown.
tgt still shows the old connections:
$ sudo tgtadm --op show --mode target
Target 5: iqn.2010-10.org.openstack:volume-b32d317e-4063-4e4f-963e-b0fdc29198f9
System information:
Driver: iscsi
State: ready
I_T nexus information:
I_T nexus: 5
Initiator: iqn.1993-08.org.debian:01:3844c047b3a9 alias:
node003-cont001
Connection: 0
IP Address: 172.21.1.3
I_T nexus: 6
Initiator: iqn.1993-08.org.debian:01:3844c047b3a9 alias:
node003-cont001
Connection: 0
IP Address: 172.21.1.3
When these duplicate connections are there, the volume cannot be
deleted:
$ sudo tgt-admin -v --force --delete
iqn.2010-10.org.openstack:volume-b32d317e-4063-4e4f-963e-b0fdc29198f9
# Removing target:
iqn.2010-10.org.openstack:volume-b32d317e-4063-4e4f-963e-b0fdc29198f9
tgtadm -C 0 --op delete --mode conn --tid 5 --sid 6 --cid 0
tgtadm -C 0 --op delete --mode conn --tid 5 --sid 5 --cid 0
tgtadm -C 0 --mode target --op delete --tid=5
tgtadm: this target is still active
Command:
tgtadm -C 0 --mode target --op delete --tid=5
exited with code: 22.
The following errors show up in syslog illustrating the force delete
failing:
Apr 9 22:31:50 node001-cont001 tgtd: conn_close_admin(237) close 6 0
Apr 9 22:31:50 node001-cont001 tgtd: tgt_event_modify(241) Cannot find event 21
Apr 9 22:31:50 node001-cont001 tgtd: iscsi_event_modify(558) tgt_event_modify
failed
Apr 9 22:31:50 node001-cont001 tgtd: conn_close_admin(237) close 5 0
Apr 9 22:31:50 node001-cont001 tgtd: tgt_event_modify(241) Cannot find event 19
Apr 9 22:31:50 node001-cont001 tgtd: iscsi_event_modify(558) tgt_event_modify
failed
Apr 9 22:31:50 node001-cont001 tgtd: tgt_target_destroy(2211) target 5 still
has it nexus
And the connections are still there:
$ sudo tgt-admin -v --show
tgtadm -C 0 --op show --mode target
Target 5: iqn.2010-10.org.openstack:volume-b32d317e-4063-4e4f-963e-b0fdc29198f9
System information:
Driver: iscsi
State: ready
I_T nexus information:
I_T nexus: 5
Initiator: iqn.1993-08.org.debian:01:3844c047b3a9 alias:
node003-cont001
Connection: 0
IP Address: 172.21.1.3
I_T nexus: 6
Initiator: iqn.1993-08.org.debian:01:3844c047b3a9 alias:
node003-cont001
Connection: 0
IP Address: 172.21.1.3
** Affects: tgt (Ubuntu)
Importance: Undecided
Status: New
** Summary changed:
- tgt connection issues can make it impossible to delete volumes
+ tgt connection issues can make it impossible to delete targets
--
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to tgt in Ubuntu.
https://bugs.launchpad.net/bugs/1305343
Title:
tgt connection issues can make it impossible to delete targets
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/tgt/+bug/1305343/+subscriptions
--
Ubuntu-server-bugs mailing list
Ubuntu-server-bugs@lists.ubuntu.com
Modify settings or unsubscribe at:
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server-bugs
