I have a Linux server running on an Asus business-class MoBo (M3N78-VM) which has on-board support for RAID. Right now I have two 1-TB drives running in a RAID-1 mirror configuration. The problem is, I don't think I have any insight into the health of the array unless I reboot and watch the BIOS screen flash by. Plus, I'd like better capability to administer the array from a shell window. I've been reading that it's better to go with either kernel-based s/w RAID, or a real h/w solution using an add-on card. So, I am thinking about converting to the s/w RAID supported by the Linux kernel. If this is a bad idea, please jump in and let me know. Otherwise, I think this is where I am headed.
My question is, can I safely convert from my existing filesystems to a kernel-based RAID setup? Using the MoBo RAID, here is how my filesystems are presented:
Code:
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/dm-2 12096756 6566112 4916160 58% /
tmpfs 1030836 0 1030836 0% /dev/shm
/dev/dm-1 101086 56300 39567 59% /boot
/dev/dm-4 960206996 469646532 490560464 49% /media
My thought was that I could simply reboot and enter the BIOS to disable RAID. Then, I am guessing that my filesystems would be presented as something like /dev/sda1, /dev/sda2, and /dev/sda3. After that, I could configure the kernel-based RAID to mirror those filesystems over to the other 1-TB drive as something like /dev/sdb1, /dev/sdb2, and /dev/sdb3. Is this feasible? Or, do I have this all wrong? I would prefer not to lose all the data I have on my existing filesystems, particularly the one mounted as /media.
Thanks!
Larry