dmesg [질문]자료첨부 했는데 다시좀 봐주세여...2006/9/14

[박선호_1]

이태곤님 답변감사드립니다 ^^



작업하 하려니 잘안되네여;;

그래서 조언을 다시구하고자 자료 첨부했습니다

자료보시구 다시한번 하는 방법 자세히 알려주셨으면 감사하겠습니다.





많은 분들의 답변 기다리겠습니다...

좋은 하루보네세여^^

[이태곤]

안녕하십니까



HBA카드나 Storage의 Controller 을 교체 하셨다면 서버을 재부팅하거나 nport ID을 replace 해주셔야 합니다.



먼저 0/0/12/1/0 가 어떤 HBA인지 확인하셔야 합니다. 여러 H/W path에서 메시지가 발생되는 것으로 보아 Storage의 Controller을 교체하신것 같은데 먼저 /dev/td 을 확인하셔야 합니다.



# ioscan -funCfc

0/0/12/1/0 가 가리키는 FC HBA을 /dev/td0 라 가정





# fcmsutil /dev/td0 devstat all | grep -e Nport -e Failed



# fcmsutil /dev/td0 replace_dsk 0x.. after exchange of a controller (0x.. = controller N_port ID)



# ioscan -f ; inef -e



=====

또는



# ioscan -fn ( NO_HW 가 있는지 확인)



NO_HW가 있다면 해당 디스크가 정상인지 확인합니다. 만약 필요없다며 rmsf 로 device 파일을 지우시면 됩니다.



참고로 다음 URL을 참고하십시오



http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=127156&admit=-682735245+1157763564915+28353475



http://www.1402.com/documents/Fibre%20Channel%20Mass%20Storage%20(FCMS).pdf#search=%22Unable%20to%20access%20previously%20accessed%20device%20at%20nport%20ID%20%22

[김병수]

관련자료입니다.



TO REPLACE A F/C DISK CONNECTED ON A F/C INTERFACE











1- Have a look in dmesg :



#dmesg



Ex:



:



:



Logical volume 64, 0x3 configured as ROOT



Logical volume 64, 0x2 configured as SWAP



Logical volume 64, 0x2 configured as DUMP



Swap device table: (start & size given in 512-byte blocks)



entry 0 - major is 64, minor is 0x2; start = 0, size = 8388608



Dump device table: (start & size given in 1-Kbyte blocks)



entry 0 - major is 31, minor is 0x16000; start = 88928, size =4194304



:



:



Memory Information:



physical page size = 4096 bytes, logical page size = 4096 bytes



Physical: 4194304 Kbytes, lockable: 3602532 Kbytes, available: 3675116



:



:



SCSI: Attempt to access partially open device -- dev: bc0b2000



SCSI: Attempt to access partially open device -- dev: bc072000



0/4/0/0: Unable to access previously accessed device at nport ID 0x90.



0/4/0/0: Unable to access previously accessed device at nport ID 0x90.



1/10/0/0: Unable to access previously accessed device at nport ID 0x90.



1/10/0/0: Unable to access previously accessed device at nport ID 0x90.







In this exemple the failing disk is at nport ID 0x90



With a double attachment, primary and alternate : 0/4/0/0 and 1/10/0/0





2- If this info doesn’t appear in dmesg try to get it from syslog.log file







3- If the disk is in a FC10 enclosure: see: FC10 loop_Id / slot table







4- Change defective disk







5- Check in ioscan :



#ioscan -fn







Class I H/W Path Driver S/W State H/W Type Description



=================================================================================



:



:



fc 0 0/4/0/0 td CLAIMED INTERFACE HP Tachyon TL/TS Fibre Channel Mass Storage Adapter



/dev/td0



:



:



target 20 0/4/0/0.8.0.255.3.2 tgt NO_HW DEVICE



disk 85 0/4/0/0.8.0.255.3.2.0 sdisk NO_HW DEVICE SEAGATE ST318451FC



/dev/dsk/c11t2d0 /dev/rdsk/c11t2d0



:



:



fc 2 1/10/0/0 td CLAIMED INTERFACE HP Tachyon TL/TS Fibre Channel Mass Storage Adapter



/dev/td2



:



:



target 64 1/10/0/0.8.0.255.3.2 tgt NO_HW DEVICE



disk 65 1/10/0/0.8.0.255.3.2.0 sdisk NO_HW DEVICE SEAGATE ST318451FC



/dev/dsk/c7t2d0 /dev/rdsk/c7t2d0







In this exemple the disk still appears as NO_HW on both attachments





6- Replace it with fcmsutil on both attachments:







# fcmsutil /dev/tdx devstat all | grep Loop (to have the corresponding N_Port IDs and Loop IDs)



#/opt/fcms/bin/fcmsutil /dev/td0 replace_dsk 0x90 (N_Port ID)



Disk at nportid 0x000090(Loop_id 50) will not be authenticated



OR



#/opt/fcms/bin/fcmsutil /dev/td0 replace dsk -l 50 (Loop ID)

(N_port ID 0x90= Loop ID 50: see table for conversion)











#/opt/fcms/bin/fcmsutil /dev/td2 replace_dsk 0x90 (N_Port ID)



Disk at nportid 0x000090(Loop_id 50) will not be authenticated



OR



#/opt/fcms/bin/fcmsutil /dev/td2 replace dsk -l 50 (Loop ID)

(N_port ID 0x90= Loop ID 50: see table for conversion)