gundersausage

The Death Of Raid

Recommended Posts

True, RAID 0 isn't more expensive if you're buying two drives anyway; however, it more or less requires you to buy two similar drives so as not to waste any capacity or performance. Running singly, you have the flexibility to, for example, get a Raptor for OS & Apps and a much bigger 7200 RPM drive for data storage, whereas RAID 0 would dictate two Raptors.

Share this post


Link to post
Share on other sites

I am re-awakening this thread due to Gilbo dissin' my banana theory. Clearly another example that the general public is not ready for this yet.

Share this post


Link to post
Share on other sites
I am re-awakening this thread due to Gilbo dissin' my banana theory.  Clearly another example that the general public is not ready for this yet.

What is your banana theory?

I may enjoy dissin' it as well.

Share this post


Link to post
Share on other sites

Here's the original appearance of the banana RAID controller. The thread includes the undeniable benchmarks.

There is also a discussion of the hardware parity calculation abilities of different bananas. Obviously if you're going to use higher RAID levels you need be aware of this information. You don't want to get stuck with poor CPU offloading abilities.

Personally, I am of the opinion that it was the lego he used --not the banana. I know I'm going to get flamed for this one though...

Share this post


Link to post
Share on other sites

The Lego? WTF.

How many different brands of Lego can you name?

Of course its the banana - its always the banana. You don't believe me?

Check here.

See? It's clearly the banana. I'm so close to the ultimate raid controller - its getting kinda tight though, the tech support for Comcast, Viewsonic, Seagate, and Whirlpool have cut me off. My last claim on my condo insurance resulted in them canceling me, i'm currently looking into litigation. Remember how Einstein was ridiculed (OMG did I spell that right?)

Enough of teaching you beginners, I have work to do.

Share this post


Link to post
Share on other sites

For the love of god! Why are you all so bent on bananas? It's ludicrous. They're fruit and FRUIT SPOILS. Do you want a halfway fluid fruit in your machine? I wouldn't.

Perhaps use nuts, they don't spoil as quickly.

Yeah, sounds good... I've got a nuts RAID 0...

Share this post


Link to post
Share on other sites

Here's what I want to know. What kind of performance do I get out of a RAID array, if I simultaneously run ALL of the preferred high level disk benchmark programs from ALL of the enthusiast sites on the Internet? :rolleyes:

Seriously folks, whatever works for you, your usage patterns and your cash flow, then by all means use it and be damned with what anyone else thinks of it.

Share this post


Link to post
Share on other sites

Am I a victim of the placebo effect? Please tell me!

I read a lot here how RAID 0 is not improving desktop performance, although it does improve the sequential transfer rate.

Well, I may be just experiencing the placebo effect, but I think it did improve performance of some apps in my case.

Here are the examples of the performance improvements I have noticed in some standard desktop applications when working with big amounts of data, tell me what you think:

1. Windows Media Player with a big media library.

Switching to the media library for the first time after player startup takes quite a lot of time with my collection of 13000 mp3s.

The library database is stored in a single 78MB file.

Seems like a sequential read to me...

2. Outlook Express startup with a lot of email (btw, i'm not using it anymore, for security reasons)

Takes a lot of time to start up with a lot of emails stored (about 2000 messages).

The email is stored in a couple of files that add up to about 300MB.

Again the read seems pretty sequential...

3. Windows Explorer

Expanding a folder the with a lot of subfolders (300+).

I think that's a sequential read, but I'm not sure. If you know NTFS, please educate me;)

If I'm correct these are examples where the sequential transfer rate plays a dominant role in slowing the application down (the drives work intensively, while the CPU doesn't) for a couple of seconds - which can be pretty irritating.

Please tell me if these examples are wrong for some reason and it's just my perception, or if these are indeed pretty much sequential reads and RAID 0 could in fact improve the performance in these cases.

Thanks!

Share this post


Link to post
Share on other sites

Another example that comes to my mind:

When I had a single drive, sometimes while defragmenting it my P2P software would break its downloads. I assumed that it couldn't get access to the drive in time to write the incoming data.

Since I got RAID 0 I have never seen it happen again, although I defrag very often.

Probably there are other explanations why this could have imporved, but I quite naturally assumed that it had to do with switching to RAID.

I see forum members here know a lot about the operation of RAID and hard drives, so perhaps you could tell me if it is even possible that RAID 0 helped or if it must have been some other factor.

Share this post


Link to post
Share on other sites

I'm wondering what kind of glue was sniffing the guy that "invented" the raid 0 + raid 1 on two partitions of two disks present in the latest Intel southbridges: doesn't the heads compete to write on one partition, then the other... If there was a benefit in going the Raid 0 way, this surely goes against it!

BUT... that solution appeared at the same time as NCQ...

So wouln't NCQ be more useful to reorder those concurrent seeks than when running disks in normal mode, or with just one partition...

Share this post


Link to post
Share on other sites

I was just searching to set up my new Epox 8kda3+ M/B with a new Raid 0 setup. (2x 200JB WD drives)

I have previously in the last few years used a few Raid 0 setups on many of computer upgrades in that time.

I first was lured to Raid 0 in all the hype & extremely high benchmarks all the review sites & forums were posting.

I myself have always thought that my setups wern't quite right, so that is why I wasn't seeing the benefits of so many others all over the net. (I put it down to me not wanting to spend all my time experimenting with different stripe sizes etc)

Now I saw another review that seem to say that the Nforce3 250GB chipset has great Raid performance. I always presumed Raid has 2 drives walking together so it got to be much faster.

My question I suppose now is am I better off putting the drives on NVidia controller as 2 seperate drives. Or mabey putting one on the NVidia controller & one on the other SATA controller so I have 1 each drive running with different drivers. (Mabey faster as their are 2 different controllers. Faster?)

I'm so glad you have opened my eyes as I kept thinking my next Raid setup HAS to be better than this.

I also have been lucky that I never had a drive fail (especially because I have always used WD drives which I have seen fail many times with others)

Like everyone I always had good intensions to backup but the capacity of the drives these days makes it quite a big process, so in the end I dont backup enough.

Sorry the long waffle but I was passing time waiting for my new Raid 0 setup to finish.

Share this post


Link to post
Share on other sites

Just one note about the mentioned Anandtech article :

Maybe it proves, that RAD-0 brings barelly any speed advantage over non-RAID, it also proves, that a Raptor brings barelly any speed advantage over an IDM DeskStar 30 GB ( while the Storage Review Performance Database claims otherwise, upto 100% performance gain with Raptor over the DeskStar ). See for example this picture :

2592.png

As you see , the RAID "advantage" is about 0.5%, while the Raptop/DeskStar advantage is 5%, while the Rator costs twice as much and is also louder.

Think about it ;-)

And to make clear : I am not making any claims about advantages or disadvantages of RAID-0, I am just pointing out the "scientificness" of the Anandtech test.

Share this post


Link to post
Share on other sites
Just one note about the mentioned Anandtech article :

Maybe it proves, that RAD-0 brings barelly any speed advantage over non-RAID, it also proves, that a Raptor brings barelly any speed advantage over an IDM DeskStar 30 GB ( while the Storage Review Performance Database claims otherwise, upto 100% performance gain with Raptor over the DeskStar ). See for example this picture :

2592.png

As you see , the RAID "advantage" is about 0.5%, while the Raptop/DeskStar advantage is 5%, while the Rator costs twice as much and is also louder.

Think about it ;-)

And to make clear : I am not making any claims about advantages or disadvantages of RAID-0, I am just pointing out the "scientificness" of the Anandtech test.

LOL; whatever you do don't bring logic to a which-hunt :D

Share this post


Link to post
Share on other sites

I've reserved comment on these proceedings because I knew there was little point to maintaining the truth in the face of a mob mentaility. But I can hold my tongue no longer after reading the crass comments in this thread towards anyone who might arrive at a different conclusion than one's own.

This is all crap. There are many problems here that invalidate the claims that RAID 0 is pointless. First off is the test rig. Guess what, the impact of disk subsystem differences is blunted when you are cpu-bound. A 2GHz P4 may be a lot more respresntative than a P3-700, but neither is totally revealing in a world of 3.5-4GHz P4s and 2.2-2.5GHz FX-53s.

Secondly, the testing regiment is crap. IPEAK results may be interesting and important in an indirect way, but they are not application testing. They attempt to model it, but are not. They prove no more than sandra drystones when it comes to comparing the effectiveness of different cpu architectures. They are the hard drive equivalent of an engine dyno, and any racer worth his salt knows you don't race dynos. The point where you arbitrarily choose what work pattern will represent this use or that use is the problem. Whatever pattern you choose is not the reality of the situation, and you have no way of quantifying if it is off or by how much for a given application.

Thirdly is the choice of controllers. Guess what, if you are going to confine yourself with a 32bit 33MHz pci controller, STR can't help much. Pretty obvious if you ask me. This would point to the advisability of using a faster bus for connection to the system. And the cost effective solution? How about ICH5/R? Hmm? Perhaps the choice of the best RAID controller for desktop use might help the desktop-oriented results... hmm.

Certainly RAID 0 degrades seek performance slightly. And that is a very important factor. But in all but the very worst case scenario the STR increase at least makes up for the loss, and in some areas allows a goodly effective increase in application performance. There is certainly the possiblitly of creating a laundry list of (cpu-limited, synthetic) benchmarks that show little advantage for it, but it is even easier to come up with a list of actual applications that do benefit.

But to do so one must buy two cutting edge drives, use the correct controller (ICH5R for desktop use), and house it in fast enough machine to allow it be of benefit. Don't judge RAID0 by its inability to transform your KT333/1GHz Tbird system, properly integrated it into a modern, optimized system design and you can indeed benefit.

And the data loss problem? The naysayers would be just as quick to nay at someone who says they can't be bothered to have implemented a proper backup regime, but then of course ignore that truth when it adds strength to the mob.

And finally is the actual experience. If you round up the latest and best components and time application load times on RAID0 vs a single identical drive, the RAID box wins. Not hugely, but it wins. Generally 5-10%. IPEAK be damned.

But I must congratulate you. The dogged pursuit of whiches real and imagined has had its impact. A fellow on my home board (ocforums.com) measured this real world performance--application load times--in a real machine (read of modern porportions):

RAID0 64k stripe [2x36GB Raptors]

UT2004......................12.12s

GC 2..........................11.81s

Lock On.....................25.96s

Far Cry....................1.23.70s

Unreal 2.....................15.90s

Single 36GB raptor

UT2004......................14.32s

GC 2..........................12.37s

Lock On.....................28.16s

Far Cry....................1.30.42s

Unreal 2.....................16.44s

Let's compare the results:

UT2004 14.32-12.12 = 2.2s

(2.2s/14.32) x 100 = 15.4% improvement

GC2 12.37-11.81 = .56s => 4.5% improvement

Lock On 28.16-25.96 = 3.2s => 11.4% improvement

Far Cry 90.42-83.70 = 6.72s => 7.5% improvement

Unreal 2 16.44-15.90 = .54s => 3.3% improvement

And even though he measured an 8.4% average improvement, immediately started chanting the anti-RAID0 mantra. I'll take my 8.4% and be happy, thank you, as this is the area least conducive to RAID0 improvement. It gets nothing but better from here.

Share this post


Link to post
Share on other sites

Huh? I don't think anyone claimed that RAID0 was slower than one drive. They said that if you already have TWO drives, then keeping them as separate drives would make it faster to copy data between the two drives than to copy things around on the RAID0. Things like WinXP bootup are unquestionably quicker with RAID0, if that is what is important to you.

Share this post


Link to post
Share on other sites

RAID0 (12 * 10K V), 2000 files, each ~12 MB:

S:\>j:dvscopy FILM2K tmp
DVS Copy FILM2K -> tmp
1900 Files 166.52 e6/s 12.51 files/s (average 167.93 e6/s) Filecount   2000
datarate    167.81 e6/s
mindatarate 33.11 e6/s
maxdatarate 245.59 e6/s
copytime    2.61 minutes

To get even close with independent disks you'd have to go SSD or RAM disk. :rolleyes:

Share this post


Link to post
Share on other sites

RAID 0 is fast, modern drives are fairly reliable. A modern home user is using much the computer to fix his photographs, edit his movies, and play awesome games which load much from the HDD... all these are uses a ''stripping" arrangement excels at.

Argh, I wish I eventually build my coveted 2x200GB Barracuda 7200.7 SATA RAID 0 array. :)

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now