SSD硬盘,先不要用在Server上

     最近突发奇想,买了两个SSD硬盘,装在DB 上,将虚拟内存及TEMP表空间,一个controlfile放在上面,希望利用SSD硬盘IO快的特点,提升DB的效能.在测试环境RUN了两星期没问题后,移到正式环境,今天中午(第3天),吃饭的路上接到同事的电话,数据库DOWN掉了.返回机房一看,放在SSD硬盘上的那个controlfile坏掉了,报如下错误：

Hex dump of (file 0, block 1) in trace file d:"mes"dump"bdump"mes_ckpt_3324.trc

Corrupt block relative dba: 0x00000001 (file 0, block 1)

Bad header found during control file header read

Data in bad block:

 type: 0 format: 0 rdba: 0x00000000

 last change scn: 0x0000.00000000 seq: 0x0 flg: 0x00

 spare1: 0x1 spare2: 0x0 spare3: 0x7c30

 consistency value in tail: 0x00000000

 check value in block header: 0x8

 block checksum disabled

Thu Sep 18 11:54:15 2008

Errors in file d:"mes"dump"bdump"mes_ckpt_3324.trc:

ORA-00202: control file: 'F:"MES"CONTROLFILE"CONTROL02.CTL'

 

Thu Sep 18 11:54:15 2008

Errors in file d:"mes"dump"bdump"mes_ckpt_3324.trc:

ORA-00227: corrupt block detected in control file: (block 1, # blocks 1)

ORA-00202: control file: 'F:"MES"CONTROLFILE"CONTROL02.CTL'

将虚拟内存移到SAN上,复制SAN上好的controlfile1到controlfile2, controlfile3,修改init文件,重新启动计算机后DB恢复正常(移动虚拟内存需要重新启动计算机).

    我喜欢尝试,时而失败.但本性难移,屡败屡战,呵呵!