So, due to a degraded HDD in my RAID5 array, I decided to build a new array.
I bought 4 320G Seagate drives… those drives have many benifits, not least of which, a 5 year warranty.
I already have a 3ware escalade RAID controller (7506-4LP), one of the best options for a hardware RAID controller.
Instead of building another RAID5, I chose to build 2, RAID1 arrays. I chose this because it’s much easier to buy 2 hard drives (upgrading one mirrored pair) than to buy 4 hard drives (to upgrade a 4 disk array). Also the second parity drive is good for data security… though, of course, if I lost 2 drives on the same array the data would still be lost.
Anyway, so I built 2 RAID1 arrays, and ended up with 2 320G paritions.
I had done some research on filesystems and wanted to move away from ext3 (due to some of the overhead and to take advantage of other filesystems benifits). I had pretty much decided on XFS, have had a previous (mildly) bad experience with ReiserFS… but while playing with my mkfs options, I noticed the overhead of the various filesystems are quite different… check out the following comparison, different empty filesystems created on the same disk and mounted:
/dev/sda1 294G 129M 279G 1% /ext3
/dev/sda1 298G 272K 298G 1% /xfs
/dev/sda1 299G 33M 299G 1% /reiserfs
Notice how little of the drive I get from ext3? It’s slower and I get less of the disk… yes, it’s compatible with ext2, but that’s not enough for me. So, I chose XFS…. I ended up just using the default mkfs.xfs options as it figured out the proper block size, etc… I was going to specify a larger log-file size, but the one it calculated was larger than the one I would have specified. I mounted the partitions with the following parameters “noatime,nodiratime,logbufs=8“.
Transferring the data back… and all seems good. I love it when things work.
— researched:
http://en.wikipedia.org/wiki/XFS
http://everything2.com/index.pl?node_id=1479435
http://terapix.iap.fr/cplt/oldSite/hard/cluster/raid_config.html