Tech Support Forum banner
Not open for further replies.
1 - 5 of 5 Posts

· Registered
1 Posts
Discussion Starter · #1 ·
Hello all,

this beeing my first post on this forum, i will make it be something really original and interesting: a blatant question about a RAID controller!

This is the situation: I've got three hard drives(40gig, 8gig, 6gig), two of which I have attached to the onboard RAID controler, and one to the IDE controller.

Currently my system boots from the IDE hard drive, which is also the smallest one, because at first i just planned to put the system only there and all the installs etc. on the other two drives. But now, I have data on all three drives, and I am noticing that the two drives that are connected to the RAID controller are working incredibly slow. Beeing the brainiac that I am, it took me a week to notice that the person who set up this system (it wasn't me, i promise!) hasn't set up an array for the two RAID drives.

My question now is, will setting up a mirror or a stripe array on the two drives that I am using currently erase all the data that is on them, or damage it in any way?

And, is there a better way to organise the drives on this system?

Take into consideration:
- Win 2000 pro
- 256 MB RAM
- The RAID is a HighPoint HPT370.
- Maxtor 5 4098U8 40GB (7200) on RAID
- IBM DTTA 350840 8 GB (3600) on RAID
- Seagate ST36422A 6 GB (3600) on IDE

Stanojevic Milan
[email protected]

Welcome to the forums Milan...................:D

The answer to your raid array question is yes. If you choose to do Raid striping 0 it will destroy any data on your dives.

As far as your setup goes I would buy another maxtor 40 gig drive to go with the one you have now and forget about the 6 and 8 gig wich are probably not ATA100 anyway which is one of the reasons it's so slow.

Save your regulade IDE channels 1 & 2 for your CD-Rom and Burner if you have one.........................:winkgrin:

· Registered
1,397 Posts
RAID stands for a Redundant Array of Inexpensive Disks. It's an old term - because disks used to be VERY expensive.

RAID has various schemes for protecting against a hard drive failing and destroying the data stored on it.

A hard disk is a mechanical device, and as such, is subject to what car owners are very familiar with - wear and tear. A Hard Drive has a life expectancy of somewhere around 4 or 5 years but some fail sooner (a lot sooner) and some haven't failed so far.

You have to understand that the heads are so close to the disk (platters) in a hard drive that a single speck of dust is too wide to fit between the head and the platter. If the head touches the platter this is called a crash.

So, anyway, someone came up with the idea of using redundant disks to store data. If one fails, you have another. Good idea.

Until recently, RAID has been the in the province of "professionals" usually because it has only been available on SCSI (Scuzzy - Small Computer Systems Interface) hard disks, not IDE hard disks which are in place in most personal computers. However, IDE RAID has made huge gains in the past few years and it is becoming more common for home users to be interested in the technology as they invest more heavily in their PC's.

Basically there are four main methods of implementing RAID.

Raid 0 is striping - meaning that 2 or more drives are written to simultaneously so data is distributed across many drives. This isn't really fault tolerant - it increases read write speed since data is read and written at higher rates because it is distributed. However, you increase your chances of suffering a hard drive failure - if one drive in a stripped array fails, the whole array is destroyed.

Raid 1 is mirroring. This is two or more drives (usually even numbers) where the same data is written to two or more drives at the same time. If one drive fails, you have a mirror copy on another drive. Depending on implementation, it is possible for the system to fail over to the mirror drive without affecting operation.

Raid 10 or 1+0 is a combination of the two above.

Raid 5 is striping with parity. This requires three or more of the SAME SIZE DRIVE. Data is written to all drives in a stripe as in RAID 0 however parity data is maintained on one "logical" drive. If one drive in this array fails, the array continues to function and the drive can be replaced and then the data can be recreated based on the parity information. Depending on implementation, it is possible to have a "hot standby" not in the array - if one fails, data is recreated on the hot standby immediately restoring fault tolerance. It is also possible to implement RAID 5 with "Hot Swappable" drives so that drives can be replaced while the system is operational without affecting performance. This is the "usual" implementation you would see in a professional IT environment.

1 - 5 of 5 Posts
Not open for further replies.