RAID

[kief]

I notice alot of people talking about thier "super fast high end systems" in this board, but one thing that strikes me as odd is noone seems to be using RAID to strip drives. Hard disks are usually the limiting factor is overall system speed nowadays since CPUs are fast and RAM is plentiful. While actually in a game it does not make a huge difference, but in loading the OS and any apps, as well as paging memory and multitasking the hard disk gets hammered. I myself run SCSI RAID in my main machine and IDE RAID on my backups and the speed is great even on the IDE. Now I realize stripped sets are not technically RAID (no redundancy, although my SCSI RAID is actually RAID as well as stripped) but the performance gain is awesome.

So to the point I was wondering if anyone else around here uses RAID and if not why and what are your opinions of it??

Jesus saves, but Mario scores!!!

 

[camieabz]

I use 2 x 40GB 60GXPs in RAID 0. Love it!

I don't worry about the redundancy stuff, as I keep twice monthly backups on CD-RWs. I also have an image of my original set-up after OS, hardware installation and ISP software, so I just replace the drive, reboot and throw on the image. One disk of files and then all the games (which don't really matter).

I don't really see the point in RAID 0+1 or RAID 1 unless you have a <b>lot</b> of critical data which would take time to restore. If its for a company or similar, then get a hot swap system.

Just as I finished typing that paragraph, I ran HD Tach. No defrag for about 5 days and 5 internet windows open (in case it matters).

Read Speed: Max = 62443.0kps Min = 17817.0kps Ave = 43186.0kps

Random Access Time = 8.3 seconds.

CPU utilisation = 12%

Read Burst Speed: Its gone off the scale, but its over 80MB per second.

<b><font color=blue>The Cisco Kid</font color=blue></b>

 

[kief]

Got a link to the bench proggie you have? I m interested to compare my IDE and SCSI setups with yours =)
I do use RAID 0/1 in one system that is semi mission critical, and has all my financial records and the like. I am far too lazy to do backups!!!
In my data machine I use RAID5 which isnt that much faster for writing, but reading REALLY smokes! It also has hot swap in case a drive dies and the controller will autorebuild from the spare in there =)

Jesus saves, but Mario scores!!!

 

[Ncogneto]

I would love to see your scores on a couple of other bench programs such as atto, and winbench 99 ( disk inspection test). Just like with andy one cpu bench mark, hdd tach does not cover the full spectrum of hard disk benchmarking and should only be used as one of several benchmarks when assesing disk performance.

It's not what they tell you, its what they don't tell you!

 

[kief]

Well id be happy to try any marks you like and see how the RAID stacks up to normal HDDs =)

Jesus saves, but Mario scores!!!

 

[svol]

I have 2 20GB ATA100 disks in RAID 0 array, unfortaintly I have the Barracuda IV drives which suck in RAID performance, but it is faster then one disk, still waiting for a solution for my slow Barracuda IV drives.

My case has so many fans that it hovers above the ground .

 

[MeTaLrOcKeR]

And why do your Barracuda's suck in Raid ????

i have a single 40GB Barracuda IV, its fast....and a buddy of mine has a couple Barracuda IV's in Raid 0 and that thing FLYS.....

-MeTaL RoCkEr

My <font color=red>Z28</font color=red> can take your <font color=blue>P4</font color=blue> off the line!

 

[FatBurger]

This has been discussed before in the hard drive section, Seagate admits that the Barracuda IV is actually slightly slower in RAID 0 than as a single drive. The drive is basically too fast, and doesn't synch up well. I can't remember the exact explanation, you should be able to search for it. There was a copy/pasted email from Seagate tech support admitting the problem.

Great drive standalone, from what I hear.

I used to run RAID 0, but the difference wasn't very noticeable, so I stopped.

<font color=orange>Quarter</font color=orange> <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>

 

[camieabz]

It would be great if everyone could post their results. I don't even know if my drives are fast, slow or average.

<A HREF="http://www.tcdlabs.com/hdtach.htm" target="_new">HD Tach</A>

<b><font color=blue>The Cisco Kid</font color=blue></b>

 

[MeTaLrOcKeR]

Hmm.....That's interesting.......

I know from personal experience there amazign in single setups........my buddies Dual setup seems to really fly too...i dunno.......mayeb its only afected with certain RAID controllers ?

-MeTaL RoCkEr

My <font color=red>Z28</font color=red> can take your <font color=blue>P4</font color=blue> off the line!

 

[svol]

Here are mine HD Tach results:

RAID 0 Array of 2 20GB Seagate Barracude IV ATA100 7200 rpm disks.

Random Acces Time: 11,1 ms
Read Burst Speed: 74,5 MB/s
Read speed: maximum 41211.0 kps, minumum 15874.0 kps, average 29311.1 kps
CPU utilization: 9.4%

My case has so many fans that it hovers above the ground .

 

[mbetea]

that is slow. pretty good burst but that won't account for much if that's the average transfer rate. i haven't done HDTach is a while, not since i have 3 60gxp's in raid0. only 2, those gave me an avg. rate of about 42mb/s-43mb/s which really isn't that much more than my single 60gxp which came in at about 39mb/s. with 3 drives in the raid now i can see a difference, really though the main reason i've kept the raid is basically because of the extra ide channels.

[insert philosophical statement here]

 

[FatBurger]

Have you actually benchmarked it? (His, not yours)
It might still be fast enough that it simply doesn't feel slow.

<font color=orange>Quarter</font color=orange> <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>

 

[AndrewT]

:frown: I get awful numbers with that test, no raid but still, even used the old and the new version just to see. <A HREF="http://andrewt1.tripod.ca/drive_test.jpg" target="_new">Screenshot of the tests</A> Drives not set to DMA, driver for Promise controller not installed but not using it. Is that normal or I got some problems? See you guys did raid tests so no clue what to compare the numbers to.

<font color=red>Handsome A7V133 looking for long term relationship with a XP CPU. Prefer non smoker.</font color=red>

 

[AndrewT]

Nevermind, it's not bad at all when have DMA on.

Maxtor 30Gig 7200Rpm ATA100 partitioned 10/20Gig
Random Access Time 10.8 ms
Read Burst Speed 75.8 Mbps
Read Speed Max 39097.0 kps
Minimum 32550.0 kps
Average 36465.7 kps
CPU Utilization 5.4 %

Seagate 40Gig 5400Rpm ATA100 partitioned 40Gig
Random Access Time 15.9 ms
Read Burst Speed 69.4 mbps
Read Speed Max 31881.0 kps
Minimum 8462.0 kps
Average 29319.8 kps
CPU Utilization 4.3 %

As usual, <A HREF="http://andrewt1.tripod.ca" target="_new">pictures</A> (when tripod isn't dead)

<font color=red>Handsome A7V133 looking for long term relationship with a XP CPU. Prefer non smoker.</font color=red>

 

[sjonnie]

Tja, why is a discussion about RAID in the CPU section? Check discussions about RAID in the HD section and noone post anything!! Is it that noone is interested in Hard Disks or that everyone just reads the CPU section and nothing else. How about a dedicated RAID section?

 

[kief]

Most peeps are in the CPU section, and the post was about raid. However it was directed at those who talk about fast CPUs making a system and the specs of the posted systems and why they dont include raid.... So indirectly it is kinda correct to be here....

Jesus saves, but Mario scores!!!

 

[kief]

I dont get good results with that program even with my SCSI raid using a $2000 scsi raid controller. I think its cuz my drive is ultra fraged right now (just did frag repot, some files are in over 2500 fragments!!! I am defragging now and will post results when its done. The seek test should be accurate even with fraged drive....6.3 ms isnt bad, but I dont think raid affects that, its just the 10k drives =)

Jesus saves, but Mario scores!!!

 

[Ncogneto]

Kief Hdd tach and SCSI don't get along well, thats why I asked you to use a couple of different Tests. Also, remember hdd tach is only performing read tests and not write tests as well. And it is only using one file size. Here is a small quote from Storage review ( the very best at hard drive reviews):

Thus, while far from perfect, WinBench 99's Disk WinMarks are among the best-available scientific, standardized approaches to drive testing. Over the last two years, we've used over 90 ATA and SCSI drives in our personal systems. We can attest through this sheer experience that performance and responsiveness as a whole correlate much more to WinBench than it does to file copies or other so-called "real world" measures.

It's not what they tell you, its what they don't tell you!

 

[AMD_Man]

I used to run RAID 0, but the difference wasn't very noticeable, so I stopped.

Not noticable? You mean in terms of overall performance or drive performance? Why would you stop though?

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor

 

[labdog]

maybe a little slow.


if you know you don't know, the way could be more easy.

 

[kief]

Hehe....I could not find winbench 99 (although I didnt look long) would you have a link to it? Or for that matter any other proggie we can all use to compare.....

Jesus saves, but Mario scores!!!

 

[kief]

Opps, nevermind, I found it with a quick search. Was looking at zd.net which is where I thought I should find it. Oh well, when I get hme from work later ill give it a try!

For anyone that wants to compare:
http://www.etestinglabs.com/benchmarks/winbench/winbench.asp

Jesus saves, but Mario scores!!!

 

[lhgpoobaa]

dont really need raid. a well optimised 60gxp and 512mb of ram does it for me.

keeps me happy

Overclocked athlon 1200C @ 8.5 x 166FSB + PC2700 = GOOD!

 

[kief]

Yeah I know for alot of stuff it wont matter. I work with video file and compression alot so that goes ALOT faster. Also I a multitaking king, and even with 1gig ram my system likes to page when I have 5 or 6 apps open, a game minimized in the background, half a dozen browser windows, nntp program, morpheus, edonkey all in the background and burning a disk at 24x. In any case you are welcome to post your optimized single drive scores so we can compare.

PS Its amazing, talk about RAID and we get good debate/info, no flames. Good thing Intel and AMD dont make raid controllers =)

Jesus saves, but Mario scores!!!