Setting up raid

[zeradul]

I just wanna know why it scores soooooooo low in pcmark 2002 I'm getting like um 506?

Because PCMARK 2002 is GARBAGE. The new 'Special Edition' WD drives are getting scores of 1400. Are they THREE times as fast as your drive? Absolutly NOT. The benchmark is Garbage.

You scored 506. 506 WHAT? And that is my point. Its a number that doesn't mean ANYTHING.

[Beefy]

Basically what zeradul said. Benchmarks don't really count for all that much. Most people just like benchmarking so they can say how good their computer is.. theoretically.. :)

[ReSpAwN DeMoN]

good point good point. so my drive is "performing" just as good as anyone elses 7200 rpm drive right? if not better cause of that faster .4ms seek time. I have cooling on my drive so I don't have to worry about the ibm drive dieing on me.

[Beefy]

forget benchmarking for a moment. Are you happy with the performance of your HDD? If you are, then don't worry about it. Leave it alone. IF it's obviously not performing as well as it should be, then start checking things out...

[ReSpAwN DeMoN]

It's fine I don't think I would even notice it if something were wrong.

[zeradul]

Benchmarks don't really count for all that much.

No! I did NOT say that ! I'm just saying that THIS benchmark score shouldn't be taken seriously, AT ALL.

As I said before, a single WD Special Edition scored 1400+ in this thread, with even some RAID 0 configs scoring less. While many single 7200 drives are scoring in the 600-700's.

So we can conclude, that this benchmark is probably testing ONLY burst speed. And since the single WD drive has 8 megs of RAM on it, It performs phenomonally in a 6 meg WRITE test. But we all know what is really happening, the ram is 1,000 times faster than the hard drive and is just caching the 6 meg file, which gives it a very inaccurate score which means, crappy benchmark.

What it comes down to this. All 'good brands' of 7200 rpm drives made in the last 2-3 years are going to perform within a small margin of each other. In fact, I would bet that margin is 25% or less. Which means, if the slowest hard drive can read at 7 megs per second, the fastest can't read more than 9 megs per second.

My suggestion to you is that if you really want to test your hard drive against your buddies drive, use HDTach. (go look for it in the HD forum) Because at least the hard drive part of PCMARK 2002 is a complete joke.

----

P.S. ATA133 will be a skipped standard. Motherboards and HD makers have announced plans to use Serial ATA 150 in the near future, with no mention of ever adopting ATA133.

[ReSpAwN DeMoN]

I'm looking for hdtach now. Is it a free utility? Also I was wondering what was happening with ata133 cause maxtor is the only company that is making ata133 drives. And I am not buying a maxtor its IBM and WD for me.

[zeradul]

Well, Maxtor is just looking for another gimic to sell drives. They first put ATA133 on their 5400 rpm drives (LOL) which of course barely see any benefit from being ATA66, Much less ATA100 or ATA133 ! Hahahaha.

[ReSpAwN DeMoN]

ata133 on a 5400rpm drive? HA what a joke.

[Beefy]

No! I did NOT say that ! I'm just saying that THIS benchmark score shouldn't be taken seriously, AT ALL.

Sorry, my bad. All I was getting at though is that as long as your system is performing OK, there's no point looking deeply into benchmark results.

[zeradul]

Beefy, I kinda knew what you were getting at, but I needed to make it clear that many Benchmarks are VERY representative of quality, capability, etc.

:thumb: :thumb:

[two5om]

ReSpAwN DeMoN

When it comes to RAIDing your drives together there are usually 2 common methods.

If you decide to go for RAID 0 then this will enable striping but remember this is without any parity which means this method will not be able to error check your data that will reside on your drives. If you wish for parity then you will have to configure your drives as RAID 5. The reason why striping is ideal for performance is that with this configuration it breaks up your data into equal pieces depending on how many drives you are using and stores each piece of data on each one of the hard drives so when it comes time for the user to recover the data it is then accessed simultaneously therefore you could say that the data recovery is practically doubled if you have RAID 0 a pair of drives.

RAID 1 is for mirroring mode which basically makes 2 copies of your data on each one of the drives. This is ideal in regards to data recovery in the case of one of your drives dying on you. For example if you are using a pair of 40GB drives which totals up to 80GB of storage you will only be able to using half of that which leaves the user with only 40GB of storage space to utilise.

I also learnt that the drives are required to be identical. Usually when I've seen people implement a RAID array they usually buy a pair of identical drives but I noticed that your drives are identical in size but not in brand name. It would be an idea to find out about that first. I think as long as they are identical in size then all should be well.

ReSpAwN DeMoN, there is a bit of configuring that goes into RAID but first you should ask yourself what sort of a configuration would you like to achieve by using RAID. Generally you'll find the following:

RAID 0 = Performance / Speed
RAID 1 = Backing up / Recovery

Each method has its advantages and disadvantages depending on what the user wishes for. Also there are approximately a total of 10 RAID arrays to choose from but you find that most of them would generally suit servers.

Keep us posted about how you go with RAIDing your drives together.... :thumb:

[Beefy]

Nice post there dude, but we've basically covered all that already.. :) Didn't want to get into too much detail with parity bit checks and the like...

[two5om]

Hehe Beefy

I had a feeling that it was a bit much but it was my very first post in the TT forums so I think I went a little all out, soz guys.... :p

I would like to say that I think that the TT forums is one of the best ones I have ever seen. What can I say, its got it all.... :D

[zeradul]

Long posts are good! They generally mean that the writer has something worth reading! And we can't have too many posters in that category! So stick around!