bootstrap$损坏模拟

--损坏前先备份system.dbf

select count(*) from bootstrap$;

delete from bootstrap$;

commit;

shutdown immediate

startup

2、通过10046定位问题

startup mount;

oradebug setmypid

oradebug tracefile_name

alter session set events '10046 trace name context forever,level 12';

alter session set db_file_multiblock_read_count=1;

alter database open;

oradebug EVENT 10046 trace name context off

从trace可以看到读取block 520到523的过程中发生报错

3、通过bbed 做块拼接

bootstrap$ 基表在相同版本相同操作系统下内容是一样的，从其他相同环境复制system.dbf过来（只需要复制前10M即可），用bbed做块复制

dd if=system01.dbf.bak of=system01_bak bs=10M count=1

copy file 10 block 520 to file 1 block 520

copy file 10 block 521 to file 1 block 521

copy file 10 block 522 to file 1 block 522

copy file 10 block 523 to file 1 block 523

sum apply