一、组网图

不涉及

二、问题描述

未按流程先在主机端清除LUN信息，直接在存储端强制取消关联LUN导致服务器端报IO error，业务无影响

三、过程分析

1、分析服务器messages日志，发现报错硬盘并未使用：

Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 Sense Key : Illegal Request [current] 
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 Add. Sense: Logical unit not supported
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 CDB: Read(10) 28 00 00 00 00 00 00 00 08 00
Mar 11 03:46:51 fybx170 kernel: blk_update_request: 310 callbacks suppressed
Mar 11 03:46:51 fybx170 kernel: blk_update_request: I/O error, dev sdty, sector 0
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 Sense Key : Illegal Request [current] 
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 Add. Sense: Logical unit not supported
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 CDB: Read(10) 28 00 0c 7f ff 80 00 00 08 00
Mar 11 03:46:51 fybx170 kernel: blk_update_request: I/O error, dev sdty, sector 209715072
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 Sense Key : Illegal Request [current] 
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 Add. Sense: Logical unit not supported
Mar 11 03:46:51 fybx170 kernel: sd 1:0:3:36: [sdty] tag#0 CDB: Read(10) 28 00 0c 7f ff f0 00 00 08 00
Mar 11 03:46:51 fybx170 kernel: blk_update_request: I/O error, dev sdty, sector 209715184

2、分析存储的odsp.log日志，发现此前有过强制取消关联LUN的操作

2018-12-07:16:23:14:0x1600001:SP2:ITL:Info:Forcibly unmap a LUN from I_T while I_T Session established successfully, <Initiator WWN>: 21:00:00:24:ff:15:8e:93, <Target port>: 50:0b:34:20:01:c8:7b:03, <LUN>: LUN-test-02.
2018-12-07:16:23:16:0x1600001:SP2:ITL:Info:Forcibly unmap a LUN from I_T while I_T Session established successfully, <Initiator WWN>: 21:00:00:24:ff:15:8e:93, <Target port>: 50:0b:34:20:01:c8:7b:03, <LUN>: LUN-test-03.
….
2018-12-07:16:26:19:0x1600001:SP2:ITL:Info:Forcibly unmap a LUN from I_T while I_T Session established successfully, <Initiator WWN>: 21:00:00:24:ff:15:8e:97, <Target port>: 50:0b:34:20:01:c8:51:02, <LUN>: LUN-test-03.
2018-12-07:16:26:22:0x1600001:SP2:ITL:Info:Forcibly unmap a LUN from I_T while I_T Session established successfully, <Initiator WWN>: 21:00:00:24:ff:15:8e:97, <Target port>: 50:0b:34:20:01:c8:51:02, <LUN>: LUN-test01.
2018-12-07:16:26:24:0x1600001:SP2:ITL:Info:Forcibly unmap a LUN from I_T while I_T Session established successfully, <Initiator WWN>: 21:00:00:24:ff:15:8e:97, <Target port>: 50:0b:34:20:01:c8:51:02, <LUN>: LUN-tset-04.

3、分析存储messages日志，发现日志中频繁有响应前端检测的报文信息

Mar 14 14:12:44 00:B3:42:01:C8:7B kernel: ODSP:MSG:TCM: TARGET_CORE: Access ACL[21:00:00:24:ff:15:8e:97] lun[33] is NULL while transport_lookup_cmd_lun
Mar 14 14:12:44 00:B3:42:01:C8:7B kernel: ODSP:MSG:TCM: TARGET_CORE: Access ACL[21:00:00:24:ff:15:8e:97] lun[33] is NULL while transport_lookup_cmd_lun
Mar 14 14:12:44 00:B3:42:01:C8:7B kernel: ODSP:MSG:TCM: TARGET_CORE: Access ACL[21:00:00:24:ff:15:8e:97] lun[33] is NULL while transport_lookup_cmd_lun

四、解决方法

1、在主机端清除离线LUN配置信息。

2、重启主机恢复。

五、风险提示

在存储端取消关联LUN之前需要按流程先在主机端清除对应的LUN注册信息。

六、关键字

强制取消关联，Lun，IO error，Access ACL，transport_lookup_cmd_lun