本站文章除注明转载外,均为本站原创: 转载自love wife & love life —Roger 的Oracle技术博客
本文链接地址: Oracle ASM crash & ORA-15196
这是一个网友的数据库,ASM数据库崩溃,但是DiskGroup无法mount了,据说是加盘导致的.
Wed Dec 10 15:44:42 GMT+08:00 2014WARNING: cache read a corrupted block gn=4 dsk=1 blk=16 from disk 1 NOTE: a corrupted block was dumped to the trace file ERROR: cache failed to read dsk=1 blk=16 from disk(s): 1 ORA-15196: invalid ASM block header [kfc.c:8281] [check_kfbh] [2147483649] [16] [2170712234 != 3074501765] System State dumped to trace file /oracle/app/admin/+ASM/bdump/+asm1_arb0_18219146.trc NOTE: cache initiating offline of disk 1 group 4 WARNING: process 18219146 initiating offline of disk 1.63799954 (JCYLDG_0001) with mask 0x3 in group 4 WARNING: Disk 1 in group 4 in mode: 0x7,state: 0x2 will be taken offline NOTE: PST update: grp = 4, dsk = 1, mode = 0x6 Wed Dec 10 15:44:42 GMT+08:00 2014ERROR: too many offline disks in PST (grp 4) Wed Dec 10 15:44:42 GMT+08:00 2014ERROR: PST-initiated MANDATORY DISMOUNT of group JCYLDG Wed Dec 10 15:44:42 GMT+08:00 2014WARNING: Disk 1 in group 4 in mode: 0x7,state: 0x2 was taken offline Wed Dec 10 15:44:42 GMT+08:00 2014NOTE: halting all I/Os to diskgroup JCYLDG!
很明显,是Oracle ASM的元数据出现一些了,从日志来看,应该是10g的库。这里我说明一点,10g的asm其实还是有点
脆弱,之前我们测试添加磁盘时ctrl+c终止SQLPLUS操作导致asm instance crash,然后再也无法mount DiskGroup的情况。
这一切在11gR2变得非常强悍了。至少目前为止我没有遇到过11gR2的ASM出现过这种情况。
针对上述的类似case恢复,其实并不是太难,在add disk的过程中Oracle ASM会更新PSU和Disk directory 元数据。
据说该环境部署了Dataguard,这样可以进行Failover切换了,这是一个好消息。
关于Oracle ASM元数据的研究,我写过一个系列的文章,大家可以参考下!如果遇到类似的故障可以联系我!
Related posts: