Re: RAID Newbie question...

TJI, late in the discussion.

Whether any raid 0, 1+0, 5, 7 or whatever will help you depends on the size of your database, raid vendor and transaction pattern.

Raid 5 (striping, parity bits, multiple drives) is very good for hardware security, particularly if the array is hot swappable. Lose a drive and don't go down.

Raid 1+0 (mirroring and more) is bettor for administration in that you can break the mirror, make a backup and restore the mirror in a 7x24 shop, while getting almost as good a security as Raid 5. Performance tends to be not as good as Raid 5.

Basically, RAIDS are for security, not performance. Remember, they are Redundant Arrays of Inexpensive Drives, designed to handle drive failure.

It is always a good idea to isolate data from indexes, both from temp and rollbacks, etc. The isolation should be at the logical drive level, without excessive concern about the physical drives. It would be better, if not best, to have each drive be a different array. If you want to go deeper, I usually consider the isolation to be at the I/O channel level, a logical drive per i/o channel.

I know a lot of people talk about striping the data in a RAID5, but I believe they are confused. If you are retrieving a large row, it makes little (some, but little) difference if that row is in 4 pieces plus parity (original RAID5) or 1 piece (RAID 0 or RAID1). Breaking a table up so that it has extents spread accross several drives, striping and balancing the data that way at the logical level, can have significant effects.

My databases range from 100GB to 1.5TB. The discussion and consideration of RAIDx works very well on the smaller databases, which have local drive arrays and disk farms. The larger databases (and the newer small ones) are using EMC and related massive storage media. EMC uses RAID internally, but the server has 1 or two links to the EMC, admittedly fiberoptic or SCON, but only 1 or two links. The number of drives, logical or physical, is irrelevant because it is all going through a minimum number of i/o channels.

Sent via Deja.com
http://www.deja.com/ Received on Wed Jan 10 2001 - 14:37:09 CST