|
This is a problem about RAID0+1. I build 2 raid0: md2(sda3,sdb3), md3(sdc3,sdd3), then build a raid1 on raid0: md4(md2,md3).
Now it can work, and I can "mke2fs -j /dev/md4" and copy files to it. But now I found a problem by doing this:
- # mount -t ext3 /dev/md4 /mnt
- # cd /mnt
- # cp /var -a ./VAR.TEST
- # dd if=/dev/zero of=./raid1on0.md4 bs=1M count=100
- # cd ..
- [b]# mke2fs -j /dev/md2 [/b]
复制代码
NOW some files in /mnt become invalide. one time be some files and the next time may be some other files.
Could I do RAID0+1? or I have done something wrong?
How can I test the raid by "destroy"?
What about the real conditions while a hard disk is broken? How can I know a disk was broken?
Thanks. |
|