[CentOS] Re: questions on kickstart

Tony Mountifield tony at softins.clara.co.uk
Fri Mar 28 21:28:25 UTC 2008


In article <47ED53AA.6040000 at pagestation.com>,
Jerry Geis <geisj at pagestation.com> wrote:
> part of my kickstart file is now:
> 
> clearpart --all --initlabel
> part --ondisk=sda raid.01 --asprimary --bytes-per-inode=4096 
> --fstype="raid" --onpart=sda1 --size=20000
> part --ondisk=sda swap    --asprimary --bytes-per-inode=4096 
> --fstype="swap" --onpart=sda2 --size=4000
> part --ondisk=sda raid.02 --asprimary --bytes-per-inode=4096 
> --fstype="raid" --onpart=sda3 --size=1     --grow
> part --ondisk=sdb raid.03 --asprimary --bytes-per-inode=4096 
> --fstype="raid" --onpart=sdb1 --size=20000
> part --ondisk=sdb swap    --asprimary --bytes-per-inode=4096 
> --fstype="swap" --onpart=sdb2 --size=4000
> part --ondisk=sdb raid.04 --asprimary --bytes-per-inode=4096 
> --fstype="raid" --onpart=sdb3 --size=1     --grow
> raid /     --bytes-per-inode=4096 --device=md0 --fstype="ext3" --level=1 
> raid.01 raid.03
> raid /home --bytes-per-inode=4096 --device=md1 --fstype="ext3" --level=1 
> raid.02 raid.04

Leaving aside the ondisk/onpart issue which has been answered, I see you
are keeping the swap partitions as individual devices.

I prefer to raid1 the swap as well, just like the filesystems. If you lose
a disk and some memory is swapped out to the failed disk, your system will
very likely crash. If swap is mirrored too, the disk failure should not
crash the system, since the same data is on the working disk.

Cheers
Tony
-- 
Tony Mountifield
Work: tony at softins.co.uk - http://www.softins.co.uk
Play: tony at mountifield.org - http://tony.mountifield.org



More information about the CentOS mailing list