Raid 0 vs 2 separate HD.

[BoberFett]

will it be faster than using just the initial 1 HD?? yes.

I call bullsh!t. Let's see some benchmarks with a single hard drive and your setup.

 

[BoberFett]

And how on earth can you figure that mounting a second volume as a directory under the first volume would give a speed advantage but setting them both up as two separate drives would not? That makes no sense.

 

[chizow]

Originally posted by: LeeTJ

and that's where we disagree. Raid 0 does not give MUCH greater performance advantage. read the article i posted.

I read that article when it was first posted, in fact, I'm quite surprised it took you ~50 posts before referencing it. There's a difference between reading it and understanding it, as those who are interested in RAID know exactly what benefits it will provide with the associated risk involved. That summary is also somewhat flawed, as it assumes cost of ownership is double, when in reality it isn't if the user intended on having 2 disks anyways.

Those numbers don't mean anything unless you provide a frame of reference. RAID 0 may provide little benefit in a task that normally takes 2 seconds, like your vanilla office program, but the task is insignificant to begin with. This has been mentioned in this thread numerous times. However, if you are loading a file that takes 2 minutes on a single IDE drive and RAID 0 cuts that time in half, those numbers mean quite a bit more don't they? Even if those type of actions only occur 5% of the time you use your PC, the point is, the benefit is very significant and real. If you do any type of sustained multimedia editing, the performance improvement is can be exponential. Furthermore, RAID 0 provides the extra storage that allows increased flexibility when working with massive multimedia files.

In any case, you'd still be better off with RAID 1, as I really don't see any performance benefit to be gained from your current implementation, and your manual backup process to external media would negate any performance increase, real or perceived.

Chiz

 

[LeeTJ]

Originally posted by: BoberFett
And how on earth can you figure that mounting a second volume as a directory under the first volume would give a speed advantage but setting them both up as two separate drives would not? That makes no sense.

you must not have read my post. i clearly said that it DIDN'T give it a performance advantage over 2 separate HD's. Where are you reading where i said that it did?

 

[BoberFett]

Originally posted by: LeeTJ

Originally posted by: BoberFett
And how on earth can you figure that mounting a second volume as a directory under the first volume would give a speed advantage but setting them both up as two separate drives would not? That makes no sense.

you must not have read my post. i clearly said that it DIDN'T give it a performance advantage over 2 separate HD's. Where are you reading where i said that it did?

I read it, I just didn't comprehend. I see what you're saying now.

 

[Goosemaster]

Any values stated are guesses at best....just to make the point.



1. Take a situation where you have two computers. Computer 1 has 2 drives with one running the OS and the other one with all the programs etc. on DIFFERENT IDE channels. Computer 2 has the drives in a RAID 0 configuration on seperate channels.


RAID DRIVE:
1. You are transmitting half of the data to 1 drive, so in sustained transfers, the drives excel as they can ideally pump oh lets say 40MB/s
so that means they can pump out 40MB/s of halves which equates to 80MB/s. now I am not stupid so I know that that will never happen.
From the immense overhead that IDE has built-in to error correction etc, these ideals speeds aren't what a consumer should expect.
Basically however, you do get faster speeds as each HD is doing a FRACTION of the work, which equated to faster speeds. NOW, acess times are different. THe drives themselves have limits, so accessing data will alow you to theorectially have data arrive in larger chuncks but still at the same speed. THink of it as how electricity works. No matter how many electrons need to be pushed through a cmedium, they will only arrive at the speed that they are pushed at, which is measured in volts.

2 SEPERATE DRIVES:
1. Lets say you load a program..."Lazy Larry's Nudist Extravaganza v. 7.6a " When the system calls for the data, it will be calling on the storage drive, and it will output it as fast as it can ideally, which as i said is 40MB/s. So it is limited by the drives physical speed. Unlike in a RAID array in cannot share the burden with another drive. THE ADVANTAGE HOWEVER is that processees on either drive will occur independantly so an OS function will have a minimal effecton a seperate programs operation if they are using solethe disk



Of course this is all ideal but just something tothink about.

 

[Smilin]

Originally posted by: BoberFett

will it be faster than using just the initial 1 HD?? yes.

I call bullsh!t. Let's see some benchmarks with a single hard drive and your setup.

Hehe you can't. I already called bullsh1t on the first page of this thread.

Someone tell me if I have this:
LEETJ wan't to put windows on C: and docs & settings & programs on D: or some variation. The idea being to let windows still read and write while an app is busy loading on a different drive, thereby eliminating the bottleneck of ATA's inability to read/write to one drive simultaneously.

YES, this would speed things up under very limited circumstances. Some other things that you may try along the same vein: Tell the OS not to page the executive, Move the pagefile to a separate partition (preferably something near the outside of the drive).

All and all I think this solution wouldn't really speed things up a ton. It's not a downright terrible idea at all but getting back to the original topic: Raid 0 vs 2 separate HD...the Raid 0 would win.