• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

raid 0

L

latief75

Limp Gawd
Joined
Mar 30, 2005
Messages
459
is raid 0 reliable ?

i have a via basied SATA controller ....8237 southbridge , ......

can anyone give me in breif steps how to setup a raid 0 array ?


thanks,
 
Reliable? Are your hard drives reliable?

You run twice the risk of losing your data. If one disk goes, you lose all your data.
 
Is there a point to Raid 0 ??

I have read in numerous places that Raid 0 doesn't really yield any performance benefits, yet it seems like a popular thing to do. So, are these reports wrong, or are people just playing around with the setups?
 
it is faster. i can feel it on my 2 raid systems

talon controller with 2 74gb raptors....install of windows in about 15minutes flat

the i have a sk43g with 2 seagate 80gb sata hdd's and its freakin fast too...

these are gaming systems, and run quite well....havnt had a problem with reliabality of data.
 
takes me 15 minutes to install windows on a single raptor....my cdrom drive slows me down more than the harddrive does. Hell, it used to take me 15 minutes with my old 80gig WD 8meg cache drive.....


Performance wise, raid 0 was designed for large, contiguous file work, like massive images or working with large video files. For random access and std home use it's show not to offer any real benifits.
 
When I had had my Raptors Raided-0 things felt faster (or maybe it was in my head). Iono NOw I just Use 1 Raptor as a Primary Drive Windows/Apps/Games and use the Second Raptor for Dowloads/Mp3's :D What a Expensive storage drive.. :)

I stripped my Raid0 in 16k seemed to have best performance in a few benchmarks I seen..
 
A) + RAID-0 will make you look l337 when you blow away your buddies in HDTach
B) + It will also actually increase performance in some tasks that stress linear transfer performance
C) - It doubles the cost of your storage subsystem while offering no where near double performance
D) - Reduced reliability sucks
E) - Despite the higher risk of failure and higher cost, it doesn't do much for most desktop usage models
 
A) + RAID-0 will make you look l337 when you blow away your buddies in HDTach
B) + It will also actually increase performance in some tasks that stress linear transfer performance
C) - It doubles the cost of your storage subsystem while offering no where near double performance
D) - Reduced reliability sucks
E) - Despite the higher risk of failure and higher cost, it doesn't do much for most desktop usage models


and hes prettty much 100% on about all that.

but it sure makes ur huevos feel big
 
Just buy 4 and go RAID 10. No more reliability problem :D

fwiw, I've never had a problem with RAID 0. For that matter, in thirteen years and across maybe twenty hard drives, I've never had a drive fail.
 
It amazes me that this is still even debated....even by the people who say it just "feels" faster. Can we please stop this needless debate anymore.

RAID0 offers ZERO performance gains, unless your using huge single files. Even then, the performance gain is only a few percentage points. Random access (common for a desktop) is not enhanced in any way, and in some cases, can even be decreased.

You do significantly increase your risk of data loss however. Theoretically, it can be argued that you cut your MTBF in half. Whatever the formula, you DO have a signifcant increase in possible data loss.

Those are facts, proven, beaten to death, and proven again. Debate if you will, but arguing against a proven fact is somewhat foolish, and definitely a waste of time.
 
djnes said:
You do significantly increase your risk of data loss however. Theoretically, it can be argued that you cut your MTBF in half. Whatever the formula, you DO have a signifcant increase in possible data loss.

Those are facts, proven, beaten to death, and proven again. Debate if you will, but arguing against a proven fact is somewhat foolish, and definitely a waste of time.
Click to expand...

I have never been able to swallow this. I guess I look at it from a different perspective. You risk of data loss with Raid0 is EXACTLY the same as if you were using a single disk. In either case, if you lose one disk you lose everything.

MTBF?
One Disk = 1,000,000 hours (whatever, I'm just pulling a number)

Raid0 Disks = 1,000,000 hours
1,000,000 hours

Your mean time before failure dosen't become 500,000 using Raid0. And it dosen't become 2,000,000 either. It is 1,000,000 hours either way. So don't let that argument push you away from RAID. And everyone else is right, performance gains aren't there.
 
upriverpaddler said:
I have never been able to swallow this. I guess I look at it from a different perspective. You risk of data loss with Raid0 is EXACTLY the same as if you were using a single disk. In either case, if you lose one disk you lose everything.

MTBF?
One Disk = 1,000,000 hours (whatever, I'm just pulling a number)

Raid0 Disks = 1,000,000 hours
1,000,000 hours

Your mean time before failure dosen't become 500,000 using Raid0. And it dosen't become 2,000,000 either. It is 1,000,000 hours either way. So don't let that argument push you away from RAID. And everyone else is right, performance gains aren't there.
Click to expand...
It's as simple as can be. In RAID0, if one drive dies, you lose the data on both drives. In comparison, RAID1 cuts your risk in half. If one drive dies, you lose nothing at all.

If your only using one drive to store your data, your relying on the integrity of one drive. If your using RAID0, your relying on the integrity of two drives, therefore, making it twice as risky. Maybe I'm not explaining it as clear as possible, but it's just basic math.
 
Yes, Raid1 does cut your risk over Raid0 and Single Disk options. But the risk with Raid0 or Single Disk is exactly the same. Whos to say you would get the top one or the bottom one when you buy just one?
 
upriverpaddler said:
But the risk with Raid0 or Single Disk is exactly the same. Whos to say you would get the top one or the bottom one when you buy just one?
Click to expand...
Your missing the point here. Your relying on two disks as opposed to one. The more disks you have, the more you have a chance of getting a bad one, or one that will die. Auto makers list their statistics of reliability in terms of defects per 1000 cars. Out of 1000 cars, some are bound to be bad (lemons). It's the same concept with hard drives. The bigger the sample, the greater chance of having one fail. Simple probability and statistics.

Now, when your talking RAID0, if one single drive fails, you lose your data. There's no redundancy. If you have 2 drives, you need both to be functioning 100% or else your data is gone.

Again, I have no idea how to make this any simpler, but it falls under the category of simple probability and statistics. Anytime you deploy a hard drive, your taking that chance that it could at some point fail. When your data needs to rely on two drives to exist, your chance just doubled of data loss. That is what MTBF is all about...predicting drive failure. When you rely on two drives, your increasing your chances.
 
Let's say that 'x' is the probability that a given hard drive will fail during its operating life.

As a single drive, the probability of complete data loss is x, naturally.
In a RAID 0 configuration, the probability of complete data loss (of the array) is 2x.
 
xonik said:
Let's say that 'x' is the probability that a given hard drive will fail during its operating life.

As a single drive, the probability of complete data loss is x, naturally.
In a RAID 0 configuration, the probability of complete data loss (of the array) is 2x.
Click to expand...
I should have used the algebra explanation. It is much simpler than my debacle.
 
Oh yeah, note that the 2x figure refers to a 2-drive RAID 0 configuration. This probability is actually nx, where 'n' is the number of drives in the array.
 
djnes said:
It amazes me that this is still even debated....even by the people who say it just "feels" faster. Can we please stop this needless debate anymore.

RAID0 offers ZERO performance gains, unless your using huge single files. Even then, the performance gain is only a few percentage points. Random access (common for a desktop) is not enhanced in any way, and in some cases, can even be decreased.

You do significantly increase your risk of data loss however. Theoretically, it can be argued that you cut your MTBF in half. Whatever the formula, you DO have a signifcant increase in possible data loss.

Those are facts, proven, beaten to death, and proven again. Debate if you will, but arguing against a proven fact is somewhat foolish, and definitely a waste of time.
Click to expand...

It does feel a lot faster. It IS faster in game loading and general tasks, especially video editing in Premiere.

How much faster...some people don't think its worth it. I like it. My roomate has the same specs as my computer, but no RAID 0, and I do load games about 2/5 faster than he does. However, a LOT depends on being able to take advantage of the speed. You need low-latency memory, and a lot of it to reduce swapping.

I get around the "twice as likely to fail" dilema by doing weekly ghost backups of the array :)
 
RAID is not going to give increases in gaming performance because the associate files aren't big enough to take advantage of it, and the same goes for general tasks such as web surfing, listening to music, watching movies, word processing, etc.

An increase will only be seen when working with large files in high-end productivity, such as video and audio encoding and editing.
 
AtomicFire said:
It does feel a lot faster. It IS faster in game loading and general tasks, especially video editing in Premiere.
Click to expand...
As others on the forum call it, this is pure FUD. Your faster game load times can be attributed to many things, but I assure you, your RAID array is not one of them. If your working with some huge files in Premiere, like files several GB in size, then I would agree you'll see a small performance gain. I'm not sure what people want, and how much more it needs to be shoved in their face, but if you read any fo the benchmark comparisons, like Anandtech's, that debunk RAID0, game load times are always part of the test bed.
 
Original Poster-
I have the same RAID controller on my KV8 Pro I believe. Most people here say it does not give any performance increace. I beg to differ, drawing upon personal experiences:

I have a pair of Seagate 7200.7 200gb SATA's. Before sticking with RAID-0 I decided to mess around. I installed an OS on one disk, no RAID array. It took me 25-30 minutes. Load times off of this fresh install were about 30-40 seconds. Quite nice.

I repeated with RAID-0 across both drives, fresh format. Install took 15-18 minutes. Boots in 15-20 seconds.

Average read rate before RAID-0: 55mb/s. After RAID-0: 96mb/s. There's an improvement.

In real-world apps, I work with 3dsMax and render 640x480 movies which are uncompressed images. 10 second clips end up being 400mb. On a single non-RAID I could not view the movies without stagger off first try, before they were cached into memory. With RAID-0, I could view them smoothly.

Games definitly load faster, also tried this.

Don't be fooled by people saying it's not good. The only backside is that you now have twice the probability of hard drive failure on the same set of data. But be wise. I suggest, RAID IT.
 
And let the FUD-spreading continue! Someone please e-mail Anand at Anandtech, and all of the other reputable sites, along with all the print mags, like MaxPC that did the actual testing and found it DOES NOT give you any performance increase. This is unbelievable how many people still buy into a myth, despite the fact it's been debunked repeatedly. Some of you still believe in Santa Claus too, right? Tooth Fairy? :rolleyes:
 
I obviously have a raid-0, and I honestly can't say if it speeds general use/windows boot/game loads up or not. If it does, I don't notice a difference. I used a single drive for a while and upgraded to two later on. Any speed increases were no more than what a fresh XP install will get you. However, I do like the convenience of 1 74GB drive C: rather than 2 36GB drives. In addition, for the occasional video or photo edit, where raid-0 has been proven to help, I like the speed bump. It's also nice to know that, even if there is little advantage to speak of, at least I'm getting the full advantage. Oh, PS, I don't store any important files on that computer long term. If, for example, I load a video or pics from my cameras and edit them, once I'm finished with them they get dumped onto my other PC's on my network. All docs are on the file store as well. If one of my drives craps, all I really lose is savegames.
 
djnes said:
And let the FUD-spreading continue! Someone please e-mail Anand at Anandtech, and all of the other reputable sites, along with all the print mags, like MaxPC that did the actual testing and found it DOES NOT give you any performance increase. This is unbelievable how many people still buy into a myth, despite the fact it's been debunked repeatedly. Some of you still believe in Santa Claus too, right? Tooth Fairy? :rolleyes:
Click to expand...
While I am one of RAID-0's biggest critics, it does improve performance in disk-disk-tape systems, is a boon to content developers, and achieves the absolute fastest file transfers possible for users that frequently do large file systems tasks. Unfortuantely, most of the users that want help with RAID-0 setups, especially here on the [H], are only out to make their buddies jealous and still cling to the myth that it offers some sort of across the board performance boost.

There are certain games (Doom 3 and WoW come to mind) that store their data in a handful of huge files. Over 90% of WoW's data is stored in eight MPQ files ranging in size from 68MB to right at 1GB. These games may benefit from RAID-0, depending on how they access those files. Even so, games that do exhibit those somewhat more linear access patters receive only mild to moderate gains from RAID-0, as a good portion of loading time is still taken up with decompression, caching of textures/models/sounds.etc, and deciding what to do next. During these times, the disk(s) may well sit idle.

RAID-0, much like SMP, is little to no benefit to desktop users. If game and application developers cleaned up the storage of the data in to larger files, they might notice RAID-0 scaling, but it would potentially come at the expense of SLED users, which make up a far larger installed base. Unfortuantely, you still have people throwing down mega bucks for SMP gaming rigs, despite that about double the cost does not net anywhere near double the performance for their application. Same idea with RAID-0.
 
DougLite said:
Unfortuantely, you still have people throwing down mega bucks for SMP gaming rigs, despite that about double the cost does not net anywhere near double the performance for their application. Same idea with RAID-0.
Click to expand...
QFT.
Except it's a bit different with RAID-0 because you are still getting 2x the storage space, as opposed to the 2nd processor in a SMP gaming system, which largely goes unused ATM. So, as I said in my above post, if you are going to add a second HD anyway, and don't have mission-critical data on your drives, RAID-0 isn't a bad way to go.
 
jebo_4jc said:
QFT.
Except it's a bit different with RAID-0 because you are still getting 2x the storage space, as opposed to the 2nd processor in a SMP gaming system, which largely goes unused ATM. So, as I said in my above post, if you are going to add a second HD anyway, and don't have mission-critical data on your drives, RAID-0 isn't a bad way to go.
Click to expand...
There's also a lot you can do with a second physical drive that's not part of the array, such as keeping a Ghost image at the ready, etc. Your data is also separated from your OS and apps, meaning it's easier to rebuild your machine as well. I could make many more arguments in favor of a second physical disk in a non-array situation. The argument about disk space is null and void, because if you separated the two disks, you'd still have the same amount of space.
 
djnes said:
And let the FUD-spreading continue! Someone please e-mail Anand at Anandtech, and all of the other reputable sites, along with all the print mags, like MaxPC that did the actual testing and found it DOES NOT give you any performance increase. This is unbelievable how many people still buy into a myth, despite the fact it's been debunked repeatedly. Some of you still believe in Santa Claus too, right? Tooth Fairy? :rolleyes:
Click to expand...

Okay. I guess my load times and renders were a figment of my imagination.
 
krizzle said:
Okay. I guess my load times and renders were a figment of my imagination.
Click to expand...
They were unscientific at best. Compared to well known sites that did actual testing, with results posted that can be repeated (the very definition of a scientific test) then yes. I can even tell you on my own that I did many a test comparing my two Raptors. Seek times go up when your using RAID0, which translates into no performance gain at all in a desktop environment. The truth of the matter is, in real world testing, RAID0 doesn't yield any performance gains. It is a known, proven fact. Whether you want to accept it or not doesn't change the fact that no performance is gained.

As far as why it "seems" faster, google "Placebo Effect".
 
I read through this thread and this one also http://www.hardforum.com/showthread.php?t=874828

And I was just curious about the article below.

http://www.viaarena.com/default.aspx?PageID=5&ArticleID=392&P=1

My interests, in the article, is with the MMO finding (RAID 0, 2x Raptors). In the article above, the writer mention that the MMO was at blazing load speeds due to the decrease in texture load times, in heavily populated areas.

I'm not sure if there's a difference in a MMO vs a FPS in performance, in regards to a RAID 0? Of course, most feedback indicates no noticeable worthy performance to use a RAID 0 for gaming. But, I was just curious.. is it for all games? or certain types?

However, there was no mention on performance with a single Raptor Drive in a MMO setup. Any opinions on this?
 
vox87 said:
I read through this thread and this one also http://www.hardforum.com/showthread.php?t=874828

And I was just curious about the article below.

http://www.viaarena.com/default.aspx?PageID=5&ArticleID=392&P=1

My interests, in the article, is with the MMO finding (RAID 0, 2x Raptors). In the article above, the writer mention that the MMO was at blazing load speeds due to the decrease in texture load times, in heavily populated areas.

I'm not sure if there's a difference in a MMO vs a FPS in performance, in regards to a RAID 0? Of course, most feedback indicates no noticeable worthy performance to use a RAID 0 for gaming. But, I was just curious.. is it for all games? or certain types?

However, there was no mention on performance with a single Raptor Drive in a MMO setup. Any opinions on this?
Click to expand...
I pointed to a potential reason for RAID-0 improving performance in certain games, and MMORPGs in particular:
DougLite said:
There are certain games (Doom 3 and WoW come to mind) that store their data in a handful of huge files. Over 90% of WoW's data is stored in eight MPQ files ranging in size from 68MB to right at 1GB. These games may benefit from RAID-0, depending on how they access those files.
Click to expand...
I do not have EQ2, so I cannot comment on how the games data is stored and accessed. Would someone kindly search for *.* in their EQ2 directory and tell me what it finds? Also, about the Via arena article, you should remember some things.

A) The drives "were kindly supplied by Western Digital"
B) They have absolutely no documented data on their conclusions about the increased MMO performance, only their observations
C) They do provide graphs of the synthetic test results (which RAID-0 wins of course)
D) They are strangely silent on performance increases in other types of games (probably because they don't exist ;))
E) They have no data from WD's competitors. A Maxtor DM10 or Hitachi (T)7K250 would have taken the 250GB WD drives to school.
F) They make next to no mention of RAID-0's huge drawbacks, doubled cost and reduced reliability.

If your WD, wouldn't it be great if you could sell two Raptors to everybody instead of merely one? Wouldn't their marketing people be absolutely giddy if they could get a review into circulation about how ninja l33t you'll be in EQ2 and that you've killed everybody and looted their bodies before they even finish loading because you have RAID-0, and WD has doubled profit on your rig?
 
K load times increase, yes. But RAID-0 is still quite nice when working with massive files, e.g. 3dSMax renders. In that respect, I'm quite happy, it was an improvement over my older non-RAID setup.
 
krizzle said:
K load times increase, yes. But RAID-0 is still quite nice when working with massive files, e.g. 3dSMax renders. In that respect, I'm quite happy, it was an improvement over my older non-RAID setup.
Click to expand...
Assuming those files are big enough, then yes, you will see a little improvement. 99.9% of the other people using a desktop PC aren't going to see that. As someone above said, they do video editing....and they do see a small increase in performance. That same person backs up their data off the array immediately, so no risk. In very unique situations RAID0 can help out a little. We're just trying to give info for the masses here.
 
You should know that a many of us on this forum falls under the 0.1% of users category. Perhaps you could be more sensitive of this, and realize that "the masses" don't frequent this forum anyways.
 
There's also a lot you can do with a second physical drive that's not part of the array, such as keeping a Ghost image at the ready, etc. Your data is also separated from your OS and apps, meaning it's easier to rebuild your machine as well. I could make many more arguments in favor of a second physical disk in a non-array situation. The argument about disk space is null and void, because if you separated the two disks, you'd still have the same amount of space.
Click to expand...

You can keep OS, apps, and data separate also by using multiple partitions on a single drive. Just mentioning it.
 
djnes said:
There's also a lot you can do with a second physical drive that's not part of the array, such as keeping a Ghost image at the ready, etc. Your data is also separated from your OS and apps, meaning it's easier to rebuild your machine as well. I could make many more arguments in favor of a second physical disk in a non-array situation. The argument about disk space is null and void, because if you separated the two disks, you'd still have the same amount of space.
Click to expand...
All 100% true, assuming the debate is 2 drives in raid vs 2 drives jbod. I guess Doug's comments about "twice the cost" etc made me think the discussion was about buying 2 drives for RAID vs only using one original drive. It doesn't matter, I'm just explaining what I was thinking.

On a related note, I love how these discussions always turn into "RAID is good" vs "Nah, RAID sux" technical debates, when the OP likely has ditched his PC all together and bought a powerbook or something by now :p
 
Quite honsetly I haven't noticed any preformance increase like I was lead to believe 2 months ago. The only actual preformance increase I noticed was reformatting and install XP pro again and again. I had one SATA WD 80gb before, so I just added another one and used the on-board ICH5 RAID on my ABIT IC7-G. It used to take me an hour or so to install XP, now it's only like 15min. Not scientific, but the only thing that changed on the system was the addition of a second physical drive and the actual array. And I reformat and install quite often.
 
I know im new here, and I hate to bring an old thread back, but..IMHO if I have 2 drives im going to use Raid0, it certainly isnt going to hurt. As far as poeple saying double the failure, a bad drives a bad drive, doesnt matter how I have it hooked up. My data is always backed up on DVD's anyhow,which everyone should be doing whether its a gaming machine,server, or whatever your using it for. Sure you dont have an increase in seek time, but reads and writes you do. There have been enough Raid0 users over the years that have reaped the benefits to prove it works. Alot of people tout articles from Storage review or Anandtech, great articles, but flawed IMO. I use it and luv it. Heres another article some people should read......

Tweakers.net Raid0 Article
 
Nice article. Whowuddathunkit? Sometimes words of Anand Tech and those "well known sites that did actual testing" aren't always correct. Not that Tweakers.net is... but they clearly sho that there IS more advantage to raid-0 than some people here choose to believe. I guess if Anand says it, it must be true, eh?
 
You must log in or register to reply here.
Back
Top