New Posts  All Forums:Forum Nav:

Raid 101

post #1 of 80
Thread Starter 
Hi Guys!

I'll be soon ordering a loaded Alienware Area-51m 7700 Extreme Notebook (Mobile Desktop)... These systems are screaming!!!!

However, I need some input please...

Not sure what is RAID? And, how are mobile hard disk drives w/ RAID configurations (e.g., RAID 0 - RAID 1) better or faster than single drive configurations? (Wouldn't a single hard drive be faster than jumping around between two separate disks, right?)

Also, any guess if Alienware will be soon carrying versions that contain Intel® Pentium® 4 Processor 570 w/ HT Technology 3.8GHz 1MB Cache -or- Intel® Pentium® 4 Extreme Edition w/ HT Technology?

What cpu is best? I was trying to compare these different chips from Intel.com (see link below)...
http://indigo.intel.com/compare_cpu/...spx?familyID=1

I appreciate any help - Thanks!
Can't wait to join the pack!
post #2 of 80
Welcome to the forums you will find that and much more information has already been discussed before. You just got to use the search function of the forum. Again welcome.

post #3 of 80
Raid is Redundant Array of Independent (or Inexpensive) Disks There are various Raid technologies. A raid can be implemented by a controller or through Windows (A controller is far faster compared being implemented by Windows) The 7700 supports Raid 0 and 1 with an onboard controller.

Raid 0 is known as striping, It takes two or more drives and makes them into 1 big drive. In Windows explorer you will see only 1 drive and it will be the capacity of both of them put together. There is a performance increase with raid 0 since the drives do the work that 1 drive would have to do.

Raid 1 is known as mirroring. You will see only one of the drives in explorer. What it does is raid 1 copies the data off of the Primary drive and places it on the 2nd drive. You will have a backup of the data and is very practical if you are saving very sensitive information and can't afford the data being lost if the drive fails.

Raid 0 - bigger drive, performance increase.
Raid 1 - redundancy, security.
post #4 of 80
Thread Starter 
Quote:
Originally Posted by jsd91741
Welcome to the forums you will find that and much more information has already been discussed before. You just got to use the search function of the forum. Again welcome.

Thanks!

I'm sure these questions have been asked but everyone knows how industry standards and technology changes overnight. Like most forum members, we want to hear what's the latest and greatest before throwing our life savings into a new system. Not to mention, getting new input and reviews from different members.

Besides when it comes to talking about cars or girls I never hear anyone say... Go look it up that's been discussed before!
post #5 of 80
dont get RAID anything. raid 0 offers no performance increase in real world applications, and doubles your chances of losing all of your data. just get 2 of 1 drive in a non raid form and use it like that.

i say again, raid 0 offers no performance increase in real apps.
post #6 of 80
Odd.. when using Visual Studio, Adobe Photoshop, Adobe Illustrator RAID0 makes a HUGE difference.
post #7 of 80
Quote:
Originally Posted by Imperil
Odd.. when using Visual Studio, Adobe Photoshop, Adobe Illustrator RAID0 makes a HUGE difference.

No it does not, Atsu is correct, bag the raid on the workstations. On servers it makes sense, a mirrored OS in a hot plug controller so you never loose the os over a failed drive. And a raid 5 data part for standard data.

There is no reason to use it on a performance system. Every time you write to a stripe your asking the controller to compute the write position, and and the locater bit for thata on the "RIS" (raid information sector) Thats takes cycles and lowers your performance.
post #8 of 80
Wierd thats odd, because our company benchmarked it before buying all new workstations and there was a nice savings in using RAID. Sure maybe not for little home use, but for large projects built in larger more hungry applications it makes a real nice difference.

The last project I just completed (relatively small/medium-small) there was a difference of 7 minutes on the compile time using VSNET. This is when using 100% identical hardware and software installs, except one machine runs in RAID0 and one does not.
post #9 of 80
there has 2 be some performance boost otherwise it would have never ben created
post #10 of 80
Quote:
Originally Posted by Flatulatta
there has 2 be some performance boost otherwise it would have never ben created
Nope, way back in the day of 2gb drives it was the only way to create cheap large volumes.
post #11 of 80
Quote:
Originally Posted by Imperil
Wierd thats odd, because our company benchmarked it before buying all new workstations and there was a nice savings in using RAID. Sure maybe not for little home use, but for large projects built in larger more hungry applications it makes a real nice difference.

The last project I just completed (relatively small/medium-small) there was a difference of 7 minutes on the compile time using VSNET. This is when using 100% identical hardware and software installs, except one machine runs in RAID0 and one does not.

Are we talking apples to apples here? Did you use a SCSI array or IDE, how many disks comprised each stripe? How many spindals? Spindals=speed
post #12 of 80
LOL RAID 0 does offer a increase in performance i have timed the loading of maps and many other aps with a stop watch . between using 2 drives in raid 0 each is 7200 rpm or using just 1 of them as a single . its not twice as fast but its still way faster. and if you use good hardrives like hitatchis that are meant to run full bore 24/7 the chance of loosing data is lower then using 1 single maxtor or westerndigital drive. Loading up your msn messenger or your yahoo you wont see a difference but any other app that relys on accessing the drive for large amounts of data you see a nice boost in speed..once you go raid 0 you never go back. Maybe 2 cpu's on 1 mother board dont make a pc faster either huh? ya know due to the fsb having to get its data from 2 sources and such. but what do i know i only been building computers for 10 years.
post #13 of 80
Quote:
Originally Posted by JUSTlN SANE
LOL RAID 0 does offer a increase in performance i have timed the loading of maps and many other aps with a stop watch . between using 2 drives in raid 0 each is 7200 rpm or using just 1 of them as a single . its not twice as fast but its still way faster. and if you use good hardrives like hitatchis that are meant to run full bore 24/7 the chance of loosing data is lower then using 1 single maxtor or westerndigital drive. Loading up your msn messenger or your yahoo you wont see a difference but any other app that relys on accessing the drive for large amounts of data you see a nice boost in speed..once you go raid 0 you never go back. Maybe 2 cpu's on 1 mother board dont make a pc faster either huh? ya know due to the fsb having to get its data from 2 sources and such. but what do i know i only been building computers for 10 years.

It's not post the numbers, and 15 years ago we had Mylex controllers that could hardley remain stable under any type of operation. Anytime you ask your controller to do math, e.g. parity operations it takes more cycles and more time. Raid 0 is the fastest raid set available, that's where the misnomer is. It's not faster then a direct write to a platter, it never will be. One needs to be computed the other does not. You may, in your case have been using a RAID controller e.g. HP's SmartArray 5I that has a large cach pool, versus a standard ATA 100 drive, and that would make your difference. Here me now and liste to me later, attach a single disk with a 2GB fiber/fabric connection to a system and the I/O transfer will slam a raid 0 set of three disks.

Also, two CPU's does not make any machine faster, they still both run a set speed, it only makes the machine more efficent at processing multiple instruction sets. Therefore an eight way machine can process eight pipes at say 1.3GHZ at the same time, but it's still 1.3GHZ. Morre in more out, same speed.

Chances are you were loading your maps from some sort of attached storage, like EMC, Shark, or Hitachi's Lightning line. Those are fiber backend solutions and they don't factor into a desktop IDE raid solution. Like I said before, "apples to apples" or what? And as far as years go, I know guys that have 20 years in that can't reload Windows XP, and guys with 1 year in that can design ane entire companies IT soltution. It's not how long, or how much you know. It's how you apply it. Every hear the term "paper cert" before?
post #14 of 80
Time-Pilot

I just can not wait for the people you got Raid 0 to start crying when they start loosing data because of the striping on both disk and they have a disk failure and wonder why there data was lost. It may appear to be faster and in some cases it may but if you look at RAID 0 Stripping the data is written to both hard drives by dividing the data between them. It like someone posted before on these forums. You can have dual 60 gig hard drives with 20gigs total data or a 40 gig with 20 gigs data or 80 gig with 20 gigs of data, which one is going to give you better performance. Once you hit 30% and more capacity of any hard drive is where your going to see a performance hit. Also last thought, its like comparing a 7200 RPM IDE to a 5400 SATA hard drive. Some may think a 5-10% performance increase is awsome, others may think 30% is more reasonable. I am just saying what I think is a good performance increase may not mean the same for the next guy. RAID 0 is good but rememeber to BACKUP DATA because if one hard drive goes your up the creek with out a paddle. This is a good thread to start and keep since now we are seeing more notebooks and PC's with RAID configurations.
post #15 of 80
Nope loaded maps from the hardrive. RAID 0 is definitly faster then non RAID. I build computers for a living and see it over and over again RAID 0 smokes non raid. in reading and writing time. infact i told a few of my fellow workers who also dabble in home built gaming rigs . they had a good luagh as well . maybe you could call up call for help on tech tv and have patrick explain it to you. and as far as loosing data i have never had a raid 0 set up fail on me yet . Nor has any of my customers i built systems for. Kidraver you missed my thread I said if you use raid be sure to get a hard drive meant to run fullbore 24/7 like a hitatchi then hd failure is as common as seeing bigfoot at mcdonalds.Remember since 2 hardrives are sharing the work of 1 they get half the wear for any given task. here are soem real world test results.
Test Method
1.) Create RAID 0 array in SATA bios.
2.) Format NTFS (default settings)
3.) Restore saved Ghost image to RAID array.
4.) Defrag RAID array
5.) Run Tests:
- Time a run of Norton Antivirus.
- Time a copy of a directory from "My Programs" to "Desktop". Directory is 2.83GB with 1410 files in 49 directories.
- Run a Sandra Filesystem Benchmark
- Run a HR Tach benchmark
- Run an ATTO benchmark.

Results

Norton Antivirus
4k Stripe - 7:43
8k Stripe - 7:58
16k Stripe - 7:54
32k Stripe - 7:57
64k Stripe - 7:59
128k Stripe - 7:52
Single Drive - 8:24

***It seems stripe size makes no difference on this test, and RAID 0 brings little improvement over a single drive.

Directory Copy
4k Stripe - 2:28
8k Stripe - 2:25
16k Stripe - 2:22
32k Stripe - 2:24
64k Stripe - 2:25
128k Stripe - 2:25
Single Drive - 3:51

***Here we see where RAID 0 performs. Stripe size does have some effect (although slight), with 16k showing the best performance.

Sandra
4k Stripe - 71mb/s
8k Stripe - 80mb/s
16k Stripe - 81mb/s
32k Stripe - 81mb/s
64k Stripe - 81mb/s
128k Stripe - 81mb/s
Single Drive - 44mb/s

***Again, we see RAID 0 perform well. Stripe size doesn't have much effect, beyond 16k.
post #16 of 80
Justin.... yes it may get a good performance out of RAID 0 but for me personally I am just not into it unless I have a good back up source. Guess I am just a "what-if" person and would like to be prepared. I would say if I am only into gaming, or performance, and need to have the hard drive accessed alot, yes I will get RAID 0 in a heart beat. It is just a personal decision for me at this point in time, not to say next week I will change my mind but right now a non-RAID array is my personal choice. I like the numbers you posted and they are correct. People need to understand your point that if you are going to do RAID properly you need the proper Hard Drives that can with stand that for years with out failing with a good RAID driver installed. Final thought is quality Hard Drives and setup will give you a quality RAID 0 configuration, poor hardware in a RAID 0 then hmmm... your not going to get the numbers Justin posted and maybe failed sets. One off the wall thought that just came up. Its pretty cool if you sit back and think about it that we are talking about RAID 0 in notebooks. It is just amazing how much notebooks have came along the past few years.

Justin..where do you get all the software to run your benchmarks. My 7700 is arriving wednesday and would like to benchmark it and maybe tweak some of the settings if I can
post #17 of 80
I wasnt intending to sound like i think every one should or needs raid bro. my point is against those who say there is no performance gain and or a higher risk of loosing data. the software can all be DL at www.futuremark.com. all versions have a free trial. Then all you need to do is errm well come across a key to unlock them . how ever the trials will give u a score. my best scores are with some slight tweaking but nothing to the point of instable i run my pc almost literly 24/7 as if it was a desktop . when its not gaming at home its at work with me. I could tweak out a few more points if i pushed it . but im pleased with what its posted.
post #18 of 80
It's not the fact that its RAID 0, that's not your gain. Anything that runs on SATA 150 is faster than UATA133. I have a single drive running n SATA mode thats showing faster numbers than what you posted.
post #19 of 80
Quote:
Originally Posted by Time-Pilot
It's not the fact that its RAID 0, that's not your gain. Anything that runs on SATA 150 is faster than UATA133. I have a single drive running n SATA mode thats showing faster numbers than what you posted.
Those numbers are from a SATA 80 gig 7200 rpm
post #20 of 80
i'm not sure why anyone would question a performance gain with raid0. it's obviously going to read/write faster than a single drive (assuming you're talking about the same drives). whether you need the drive speed is a different story. and i have no clue what's controlling the raid in the 7700, mine's still being built...

i wouldn't trust a single drive any more than a raid0. i don't believe in "more likely to fail", just back it up if it matters, raid0 or not.
New Posts  All Forums:Forum Nav:
  Return Home