asm 磁碟切換

babyyellow發表於2012-04-28
ASM
計劃本週做 asm 磁碟的切換準備工作，上週已經做好了lun的劃分工作，

sunha5_n 包括兩個磁碟組， sunha5_f1, sunha5_f2  ，我們原定的分別跨在兩個儲存上，

前段時間，因為磁碟空間緊張，臨時新增磁碟空間打破了這種格局，對維護帶來一定的難度，需要調整回去，本次操作的目的在於此

刪除  sunha5_n_0000 磁碟，新增一個新磁碟代替之。 

先是刪除很輕鬆

01  1* select name, failgroup ,path from v$asm_disk where name is not null order by name
02SYS AS SYSDBA at +ASM > /
03 
04NAME                 FAILGROUP       PATH
05-------------------- --------------- --------------------------------------------------
06SUNHA5_N_0000        SUNHA5_F2       /dev/rdsk/c4t6006016010511E008C4BB66773C4DF11d0s6
07SUNHA5_N_0001        SUNHA5_F2       /dev/rdsk/c4t6006016023912800BA383E8BACBADF11d0s6
08SUNHA5_N_0002        SUNHA5_F2       /dev/rdsk/c4t6006016023912800B02840888FB1DF11d0s6
09SUNHA5_N_0003        SUNHA5_F2       /dev/rdsk/c4t60060160239128008E54346EACBADF11d0s6
10SUNHA5_N_0004        SUNHA5_F1       /dev/rdsk/c4t6006016010511E00128706D3AABADF11d0s6
11SUNHA5_N_0005        SUNHA5_F1       /dev/rdsk/c4t6006016010511E002D5FFF791CB9DF11d0s6
12SUNHA5_N_0006        SUNHA5_F1       /dev/rdsk/c4t6006016010511E007C165A71C0A5DF11d0s6
13SUNHA5_N_0007        SUNHA5_F1       /dev/rdsk/c4t6006016010511E00E4F60988C0A5DF11d0s6
14SUNHA5_N_0008        SUNHA5_F2       /dev/rdsk/c4t6006016023912800CBCB861C1FB9DF11d0s6
15SUNHA5_N_0009        SUNHA5_F1       /dev/rdsk/c4t6006016010511E008E4BB66773C4DF11d0s6


1SYS AS SYSDBA at +ASM > alter diskgroup sunha5_n   drop disk SUNHA5_N_0000 ;
2 
3Diskgroup altered.


新增新磁碟 
01SYS AS SYSDBA at +ASM > alter diskgroup sunha5_n  add  failgroup sunha5_f2 disk '/dev/rdsk/c4t6006016023912800AE557B32B813E011d0s6';
02alter diskgroup sunha5_n  add  failgroup sunha5_f2 disk '/dev/rdsk/c4t6006016023912800AE557B32B813E011d0s6'
03*
04ERROR at line 1:
05ORA-15032: not all alterations performed
06ORA-15010: name is already used by an existing ASM disk
07 
08 
09SYS AS SYSDBA at +ASM > edit
10Wrote file afiedt.buf
11"afiedt.buf" 2 DD￡？110 ×？·？
12 
13alter diskgroup sunha5_n  add  failgroup sunha5_f2 disk '/dev/rdsk/c4t6006016023912800AF557B32B813E011d0s6'
14/
15~

系統報錯，無法新增磁碟 ，問題出在哪裡呢？ 

這個錯誤時報 asm   磁碟的 名 被佔用，為新加磁碟指定磁碟名，當時，無法確定是什麼問題，於是決定先回滾
01SYS AS SYSDBA at +ASM > alter diskgroup sunha5_n undrop disks;
02 
03Diskgroup altered.
04 
05  1* select name,failgroup,path from v$asm_disk order by failgroup ,name
06SYS AS SYSDBA at +ASM > /
07 
08NAME                 FAILGROUP       PATH
09-------------------- --------------- --------------------------------------------------
10SUNHA5_N_0004        SUNHA5_F1       /dev/rdsk/c4t6006016010511E00128706D3AABADF11d0s6
11SUNHA5_N_0005        SUNHA5_F1       /dev/rdsk/c4t6006016010511E002D5FFF791CB9DF11d0s6
12SUNHA5_N_0006        SUNHA5_F1       /dev/rdsk/c4t6006016010511E007C165A71C0A5DF11d0s6
13SUNHA5_N_0007        SUNHA5_F1       /dev/rdsk/c4t6006016010511E00E4F60988C0A5DF11d0s6
14SUNHA5_N_0009        SUNHA5_F1       /dev/rdsk/c4t6006016010511E008E4BB66773C4DF11d0s6
15SUNHA5_N_0000        SUNHA5_F2       /dev/rdsk/c4t6006016010511E008C4BB66773C4DF11d0s6
16SUNHA5_N_0001        SUNHA5_F2       /dev/rdsk/c4t6006016023912800BA383E8BACBADF11d0s6
17SUNHA5_N_0002        SUNHA5_F2       /dev/rdsk/c4t6006016023912800B02840888FB1DF11d0s6
18SUNHA5_N_0003        SUNHA5_F2       /dev/rdsk/c4t60060160239128008E54346EACBADF11d0s6
19SUNHA5_N_0008        SUNHA5_F2       /dev/rdsk/c4t6006016023912800CBCB861C1FB9DF11d0s6
20 
21  磁碟又找回來了


進過確認，發現新增的磁碟是新開的lun 沒有使用過，不存在什麼問題，許可權也夠，問題應該是出在asm 上

1SYS AS SYSDBA at +ASM > select * from v$asm_operation;
2 
3GROUP_NUMBER OPERATION  STATE         POWER     ACTUAL      SOFAR   EST_WORK   EST_RATE EST_MINUTES
4------------ ---------- -------- ---------- ---------- ---------- ---------- ---------- -----------
5           2 REBAL      RUN               1          1      23154      95589       1470          49

原來刪除了之後，系統自動發動了rebalance 操作，先前發出的三處的命令，並沒有真正的刪除磁碟，需要先吧這個磁碟的資料，移動到其他盤上，才會真正的刪除磁碟，而asm 分配新加磁碟的命令是指派的asm 磁碟名稱總是從最小沒有使用的磁碟名開始，
 導致  把 sunha5_n-0000  分給了新磁碟  而舊盤還沒有從cache中清理掉，導致了衝突。

知道了原因就好辦了，為了節省rebalance 的時間，我們採用了先加盤，然後再刪除盤，  有個缺點會導致當前命令視窗阻塞。 
1SYS AS SYSDBA at +ASM >alter diskgroup  sunha5_n
2add failgroup sunha5_f2 disk '/dev/rdsk/c4t6006016023912800AE557B32B813E011d0s6'
3name 'SUNHA5_N_0010' reba
4lance wait
5/


另開一個視窗 刪除磁碟sunha5_n_0000

01SYS AS SYSDBA at +ASM > col name for a20
02SYS AS SYSDBA at +ASM > col path for a50
03SYS AS SYSDBA at +ASM > col failgroup for a10
04SYS AS SYSDBA at +ASM > select name ,failgroup ,path from v$asm_disk where name is not null order by failgroup ,name;
05 
06NAME                 FAILGROUP  PATH
07-------------------- ---------- --------------------------------------------------
08SUNHA5_N_0004        SUNHA5_F1  /dev/rdsk/c4t6006016010511E00128706D3AABADF11d0s6
09SUNHA5_N_0005        SUNHA5_F1  /dev/rdsk/c4t6006016010511E002D5FFF791CB9DF11d0s6
10SUNHA5_N_0006        SUNHA5_F1  /dev/rdsk/c4t6006016010511E007C165A71C0A5DF11d0s6
11SUNHA5_N_0007        SUNHA5_F1  /dev/rdsk/c4t6006016010511E00E4F60988C0A5DF11d0s6
12SUNHA5_N_0009        SUNHA5_F1  /dev/rdsk/c4t6006016010511E008E4BB66773C4DF11d0s6
13SUNHA5_N_0000        SUNHA5_F2  /dev/rdsk/c4t6006016010511E008C4BB66773C4DF11d0s6
14SUNHA5_N_0001        SUNHA5_F2  /dev/rdsk/c4t6006016023912800BA383E8BACBADF11d0s6
15SUNHA5_N_0002        SUNHA5_F2  /dev/rdsk/c4t6006016023912800B02840888FB1DF11d0s6
16SUNHA5_N_0003        SUNHA5_F2  /dev/rdsk/c4t60060160239128008E54346EACBADF11d0s6
17SUNHA5_N_0008        SUNHA5_F2  /dev/rdsk/c4t6006016023912800CBCB861C1FB9DF11d0s6
18SUNHA5_N_0010        SUNHA5_F2  /dev/rdsk/c4t6006016023912800AE557B32B813E011d0s6
19新磁碟已經新增上去了
20 
21 
22SYS AS SYSDBA at +ASM > alter diskgroup sunha5_n drop disk SUNHA5_N_0000;
23 
24 
25Diskgroup altered.
26刪掉舊盤
27SYS AS SYSDBA at +ASM > select * from v$asm_operation ;
28 
29GROUP_NUMBER OPERATION  STATE         POWER     ACTUAL      SOFAR   EST_WORK   EST_RATE EST_MINUTES
30------------ ---------- -------- ---------- ---------- ---------- ---------- ---------- -----------
31           1 REBAL      RUN               1          1       4802     121171       1850          62
32 
33 
34SYS AS SYSDBA at +ASM > alter diskgroup sunha5_n rebalance power 5 ;
35 
36Diskgroup altered.
37並行rebalance 
38SYS AS SYSDBA at +ASM > select * from v$asm_operation ;
39 
40GROUP_NUMBER OPERATION  STATE         POWER     ACTUAL      SOFAR   EST_WORK   EST_RATE EST_MINUTES
41------------ ---------- -------- ---------- ---------- ---------- ---------- ---------- -----------
42           1 REBAL      RUN               5          5         72      86286       1993         


執行過程中，我們的asm 磁碟組檢查程式碼成功的檢測到asm磁碟故障。 


NAME                   STATE        MOUNT_STAT       PATH                                                 TOTAL(GB)   USED%                                                         
--------------------   ----------   --------------   --------------------------------------------------   ---------   ----------                                                    
SUNHA5_N_0000          DROPPING     CACHED           /dev/rdsk/c4t6006016010511E008C4BB66773C4DF11d0s6        99.75   36.8                                                          
SUNHA5_N_0001          NORMAL       CACHED           /dev/rdsk/c4t6006016023912800BA383E8BACBADF11d0s6        99.75   80.49                                                         
SUNHA5_N_0002          NORMAL       CACHED           /dev/rdsk/c4t6006016023912800B02840888FB1DF11d0s6        99.75   80.49                                                         
SUNHA5_N_0003          NORMAL       CACHED           /dev/rdsk/c4t60060160239128008E54346EACBADF11d0s6        99.75   80.5                                                          
SUNHA5_N_0004          NORMAL       CACHED           /dev/rdsk/c4t6006016010511E00128706D3AABADF11d0s6        99.75   80.49                                                         
SUNHA5_N_0005          NORMAL       CACHED           /dev/rdsk/c4t6006016010511E002D5FFF791CB9DF11d0s6        99.75   80.49                                                         
SUNHA5_N_0006          NORMAL       CACHED           /dev/rdsk/c4t6006016010511E007C165A71C0A5DF11d0s6        99.75   80.49                                                         
SUNHA5_N_0007          NORMAL       CACHED           /dev/rdsk/c4t6006016010511E00E4F60988C0A5DF11d0s6        99.75   80.49                                                         
SUNHA5_N_0008          NORMAL       CACHED           /dev/rdsk/c4t6006016023912800CBCB861C1FB9DF11d0s6        99.75   80.49                                                         
SUNHA5_N_0009          NORMAL       CACHED           /dev/rdsk/c4t6006016010511E008E4BB66773C4DF11d0s6        99.75   80.49                                                         
SUNHA5_N_0010          NORMAL       CACHED           /dev/rdsk/c4t6006016023912800AE557B32B813E011d0s6        99.75   43.69          


最後終於完成了
01  1* select name ,failgroup ,path from v$asm_disk where name is not null order by failgroup,name
02SYS AS SYSDBA at +ASM > /
03 
04NAME                 FAILGROUP  PATH
05-------------------- ---------- --------------------------------------------------
06SUNHA5_N_0004        SUNHA5_F1  /dev/rdsk/c4t6006016010511E00128706D3AABADF11d0s6
07SUNHA5_N_0005        SUNHA5_F1  /dev/rdsk/c4t6006016010511E002D5FFF791CB9DF11d0s6
08SUNHA5_N_0006        SUNHA5_F1  /dev/rdsk/c4t6006016010511E007C165A71C0A5DF11d0s6
09SUNHA5_N_0007        SUNHA5_F1  /dev/rdsk/c4t6006016010511E00E4F60988C0A5DF11d0s6
10SUNHA5_N_0009        SUNHA5_F1  /dev/rdsk/c4t6006016010511E008E4BB66773C4DF11d0s6
11SUNHA5_N_0001        SUNHA5_F2  /dev/rdsk/c4t6006016023912800BA383E8BACBADF11d0s6
12SUNHA5_N_0002        SUNHA5_F2  /dev/rdsk/c4t6006016023912800B02840888FB1DF11d0s6
13SUNHA5_N_0003        SUNHA5_F2  /dev/rdsk/c4t60060160239128008E54346EACBADF11d0s6
14SUNHA5_N_0008        SUNHA5_F2  /dev/rdsk/c4t6006016023912800CBCB861C1FB9DF11d0s6
15SUNHA5_N_0010        SUNHA5_F2  /dev/rdsk/c4t6006016023912800AE557B32B813E011d0s6

來自 “ ITPUB部落格 ” ，連結：http://blog.itpub.net/133735/viewspace-722451/，如需轉載，請註明出處，否則將追究法律責任。

相關文章