问题:
以前都备份很好的磁带备份,今天数据库无法成功备份了
环境:
数据库: oracle 10g
主机:rx7620
操作系统:hpux 11.23
带库:HP MSL6000
备份软件:hp dp 6.0
分析:
1. 在dp manager gui中,看到以下错误:
在report event中,看到以下错误:
138:707] Backup session “2017/12/27-4” of session specification “Oracle8 oracle_backup” has errors: 10.
Description:
Errors have been detected during backup.
Actions:
Check error messages of the session by using:
* Data Protector GUI (Internal Database context)
* Backup Errors report
* omnidb -session <sessionID> -report
2.通过internal database –session,看到失败的session中,以下报警:
…
[Major] From: BMA@jh7601 “HP:Ultrium 3-SCSI_1” Time: 2017-12-27 18:54:03
[90:63] By: UMA@jh7601@/dev/rac/c18t0d0
Cannot load exchanger medium ([2] No such file or directory)
[Critical] From: UMA@jh7601 “HP:MSL6000 Series” Time: 2017-12-27 18:54:28
[90:59] jh7601 : /dev/rac/c18t0d0
Cannot open exchanger control device ([2] No such file or directory)
…
3.通过 device & media –>enviroment –>devices –>HP MSL6000 Series –>robotics path
发现: /dev/rac/c18t0d0 是jh7601 的路径:
4.device & media –>enviroment –>devices –>HP MSL6000 Series 右键-> 属性 –>tab–> control –>
客户端:jh7601–> scsi address 下拉菜单,进行scsi 地址扫描,等一会,报 错误:
No robotic devices detected
159:13044
5.devbra -dev 查看带库控制台端,所连驱动器
jh7601[/]#/opt/omni/lbin/devbra -dev
Tape HP:C7438A Path: “/dev/rmt/0mn” SN: “HU10652SG9”
Description: HP StorageWorks DAT 72 Drive
Revision: ZP5A Device type: 4mm [1] Flags: 0x0001
jh7601[/]#
没有看到驱动器,正常的应该有驱动器,看正常的应该是这个:
jh7602[/]#/opt/omni/lbin/devbra -dev
Exch HP:MSL6000 Series Path: “/dev/rac/c16t0d0” SN: “USX650Z04F”
Description: HP StorageWorks MSL 6000 Series
Revision: 0518 Flags: 0x0006 Slots: 29 Drives: 2
Drive(s) SN:
“HU10728M3G”
“HU10646B54”
Tape HP:Ultrium 3-SCSI Path: “/dev/rmt/5mn” SN: “HU10728M3G”
Description: HP LTO drive
Revision: G65W Device type: lto [13] Flags: 0x0001
Tape HP:Ultrium 3-SCSI Path: “/dev/rmt/6mn” SN: “HU10646B54”
Description: HP LTO drive
Revision: G25W Device type: lto [13] Flags: 0x0001
Tape HP:C7438A Path: “/dev/rmt/0mn” SN: “HU10652SG4”
Description: HP StorageWorks DAT 72 Drive
Revision: ZP5A Device type: 4mm [1] Flags: 0x0001
6.扫描服务器上磁带设备信息:
jh7601[/]#ioscan -fnC tape
Class I H/W Path Driver S/W State H/W Type Description
=========================================================================
tape 5 0/0/10/1/0.1.7.255.0.0.1 stape NO_HW DEVICE HP Ultrium 3-SCSI
/dev/rmt/5m /dev/rmt/5mn /dev/rmt/c18t0d1BEST /dev/rmt/c18t0d1BESTn
/dev/rmt/5mb /dev/rmt/5mnb /dev/rmt/c18t0d1BESTb /dev/rmt/c18t0d1BESTnb
tape 6 0/0/10/1/0.1.7.255.0.0.2 stape NO_HW DEVICE HP Ultrium 3-SCSI
/dev/rmt/6m /dev/rmt/6mn /dev/rmt/c18t0d2BEST /dev/rmt/c18t0d2BESTn
/dev/rmt/6mb /dev/rmt/6mnb /dev/rmt/c18t0d2BESTb /dev/rmt/c18t0d2BESTnb
tape 0 1/0/0/3/1.3.0 stape CLAIMED DEVICE HP C7438A
/dev/rmt/0m /dev/rmt/0mnb /dev/rmt/c3t3d0BESTn /dev/rmt/c3t3d0DDSb
/dev/rmt/0mb /dev/rmt/c3t3d0BEST /dev/rmt/c3t3d0BESTnb /dev/rmt/c3t3d0DDSn
/dev/rmt/0mn /dev/rmt/c3t3d0BESTb /dev/rmt/c3t3d0DDS /dev/rmt/c3t3d0DDSnb
jh7601[/]#
看到磁带设备的 h/w type为 NO_HW,表明以前是存在的,现在不存在了,
正常的应该是这样:
jh7602[/]#ioscan -fnC tape
Class I H/W Path Driver S/W State H/W Type Description
=========================================================================
tape 5 0/0/10/1/0.1.7.255.0.0.1 stape CLAIMED DEVICE HP Ultrium 3-SCSI
/dev/rmt/5m /dev/rmt/5mn /dev/rmt/c16t0d1BEST /dev/rmt/c16t0d1BESTn
/dev/rmt/5mb /dev/rmt/5mnb /dev/rmt/c16t0d1BESTb /dev/rmt/c16t0d1BESTnb
tape 6 0/0/10/1/0.1.7.255.0.0.2 stape CLAIMED DEVICE HP Ultrium 3-SCSI
/dev/rmt/6m /dev/rmt/6mn /dev/rmt/c16t0d2BEST /dev/rmt/c16t0d2BESTn
/dev/rmt/6mb /dev/rmt/6mnb /dev/rmt/c16t0d2BESTb /dev/rmt/c16t0d2BESTnb
tape 0 1/0/0/3/1.3.0 stape CLAIMED DEVICE HP C7438A
/dev/rmt/0m /dev/rmt/0mnb /dev/rmt/c3t3d0BESTn /dev/rmt/c3t3d0DDSb
/dev/rmt/0mb /dev/rmt/c3t3d0BEST /dev/rmt/c3t3d0BESTnb /dev/rmt/c3t3d0DDSn
/dev/rmt/0mn /dev/rmt/c3t3d0BESTb /dev/rmt/c3t3d0DDS /dev/rmt/c3t3d0DDSnb
jh7602[/]#
结论:
基于此,初步判定,应该是出现备份问题的服务器与磁带之间的物理连接出现问题,让前端维护人员检查光迁连接线路。
已经确认,确实是主机到磁带的光迁链路出现问题,更换光迁线后,问题解决.