Slashdot Mirror
Drive Speed Comparisons - 7200RPM vs. 5400RPM in RAID?
O asks: "Looking at Pricewatch for new drives, it seems that the prices for a 60GB 5400rpm drive and a 40GB 7200rpm drive are about the same. I do have ATA/100 RAID on my motherboard, and I'm wondering how a striped RAID of 5400rpm drives would stack up against a single 7200rpm drive? I could sure get a hell of a lot more storage for the money. Any Slashdot readers have experience with this?"
The performance will depend on the application mix you are using. How much data can the drive continiously stream in/out which is the dependant on the drive and the bandwidth of the controller. Are you needs read intensive, write intensive, random or linear in nature? I would almost always recommend using hardware raid to anyone because you get reliability, which is probably more imporant anyway. A stripped RAID array will be able to provide better read throughput, especially as the number of drives goes up and the reads can be spread accross multiple heads. If you are doing random editing of high bandwidth video, however, you probobly want mirrored high RPM drives. This allows for full spead reads and writes, and redundancy. If you are doing lots of writes, especially in small chunks, stripped RAID is not going to be as fast--too much parity calculation. Also, it depends on the quality of the hardware RAID implementation, the saturation of the bus (are you going to run your ATA/100 drives as masters in independant channels?) etc.
You will probobly be better off getting something which meets your needs rather that worrying about the best performance. So my suggestion would be RAID for the added safety value (not to mention that unless you need it, more space will probobly be worth more than a few milliseconds of performance gain). Remember though, RAID doesn't save you from powerfailures! If you are going to have any stripped RAID array, even hardware, you must have it on a UPS so that it is not left in an inconsistant state! Good hunting!
In terms of throughput, yes, you'll get about the same performance. However, unless you spend all day loading large files from a completely defragmented partition, what you should really look for is reduced seek times. This is the bottleneck when doing 99% of the work in your computer -- and that's why solid state drives are so blazingly fast, despite having an inferior throughput than their mechanical counterparts: they have smaller seek times, by many orders of magnitude.
Smaller seek times can be achieved by various means:
1. Get a high-end drive. More expensive models feature faster actuators.
2. Reduce the physical area the disk can access, by partitioning accordingly. Seagate also used this trick with their first-gen X15 drive, by reducing the platter size from 3" to 2.5", and they were very successful.
3. Use RAID 1 (mirroring) with a good controller, which for IDE basically means the 3ware models. They are, far and away, the best. Besides shaving off the seek times, they also improve the throughput.
4. And, of course, the most important: get a hard drive which spins faster. If you don't believe me, take a look at the sites below for benchmarks.
The hard drive is the only peripheral nowadays whose access time is measured in miliseconds and not nanoseconds. Instead of buying loads of memory and the fastest processor available, everyone should pay attention to the storage side of their machines -- whoever has experienced 10k and 15k RPM disks finds it hard to go back.
The best places in the web for storage info are Storage Review and, to a lesser extent, x-bit labs. Storage Review, besides their extremely scientific methodologies, maintains a drive reliability database, so you can check whether the drive you're looking for is likely to fail.
Join the NFSNET. Our prime goal is making little numbers out of big ones. http://www.nfsnet.org/
We put in 10,000 RPM drives. The improvement over 5400 was definately visable just in the operating system. But I'd say 7200 should make improvements even if you don't see them. And if the prices aren't that far apart, I'd just grab the faster (if its only a few bucks more). Just my 2 copper round thingies
Can all fish swim?
Apparently, the guy who asked the question was looking to strip his drives in a RAID 0 configuration. This config won't add reliability at all, and in fact will decrease it proportionally to the amount of drives used: a single drive fails, you lose all your data. ALL OF IT, let that be clear. No going back. And this factor is increasingly more important in these days of the Deskstar 75GXP.
However, RAID 5 (which is what I believe you refer to) can indeed add reliability, but at a price: one of your disks' worth of space is ``wasted'', and drive performance can decrease along with an increase in CPU usage, unless you have a high quality controller.
Of course, RAID 0's twice as high likelihood of failure may not be all that important if you're only storing your MP3s, but keep it in mind if you're storing important data on it.
Join the NFSNET. Our prime goal is making little numbers out of big ones. http://www.nfsnet.org/
Forgot to mention in my previous comment -- this onboard RAID featured in your motherboard, whatever manufacturer it may be, is not equivalent to a real RAID board with a microprocessor inside, which offloads all the processing. Imagine Winmodem -- all processing is done on the host. Of course a modem's work isn't as CPU intensive as the RAID work, but the data rate here is of a few megabytes per second. You will end up paying with your memory bandwidth too. If you're going the RAID way, try to avoid these ``WinRAID'' controllers; buy a 3ware instead, or perhaps an Adaptec.
Join the NFSNET. Our prime goal is making little numbers out of big ones. http://www.nfsnet.org/
When striping, what's a good size to use if the goal is speeding up random access, not overall throughput?
Suppose I'm wanting to speed up random access to thousands of files, mostly 30k or less, read from or written to in full each time they're accessed. Do I make my stripe size some multiple of that 30k? Is striping even the best solution?
What about striping to increase overall throughput on large streams? Do you go with some percentage of the drive's cache size? What gives the best performance on long contiguous streams? Again, is there a better solution than striping if performance is the concern, not redundancy?
If I got three to five superfast SCSI drives, what's the best way to configure those?
That's the biggest myth out there. The faster the hard drive spins, the quicker the access time. Well, that is (for the most part) true with SCSI, it isn't with IDE. Let's take a look at 2 ATA/100 40gb drives by Seagate, shall we?
U Series (ST340823A) 5400 RPM seek time: 8.9ms
Barracuda (ST340824A) 7200 RPM seek time: 8.9ms
Wow, they are the same - so why should I go 7200 then? Simple. That 5400 RPM drive has a transfer rate of 376mbits/sec while the 7200RPM guy does 500mbits/sec. A good deal faster.
If your going to be moving a lot of big files, then the Barracuda is obviously the better choice. For a lot of small files, either one is fine. Or opt for a Cheetah 15k rpm SCSI drive, with a seek time of 3.9ms and 690mbits trans. rate.
But overall, the price between the 2 IDE drives is so slim as to be a no brainer, you can't go wrong with the 7200 rpm drive - whereas you could cripple yourself some with that 5400. So get the 7200.
Looking for hardware (Currently need: Large Etch-a-Sketch) Have one? See my journal!
I would far from quantify it from 'wasted.' It's a tradeoff. And the next level up, stripes across mirrored sets, you use 2 drives to store one drive worth of space. Basically, though, if you need to ask, you probably don't need RAID. :-)
Vintage computer games and RPG books available. Email me if you're interested.
Compaq Smart Array rock like a motherfucker. Get two, load them up with cache, and stick one on each PCI bus. Nice.
If you want to load big applications and deal with big files and have fast loading of big files raid is good.
but if you want to deal with many little small files then its better to not have raid.
I dont think you'll be dealing with many small files unless you host a server, so for most people raid is better.
If you use Linux, please help development of Autopac
Your post was good up until the part about recommending StorageReview's drive reliability database. Self-selected surveys don't reveal useful reliability statistics. What you're seeing on that database is anything but useful for checking to see if your drive is likely to fail because the only people who post to that database are those with an axe to grindthose who decided to participate themselves. On top of that there is no worthwhile way to determine if someone is posting multiple times (perhaps someone from company X griping about company Y's drive).
If the folks at StorageReview had done a random phone poll and reported the results you'd have something to look at. As it is, it's an excercise in website form submission handling more than anything else.
See any survey research or sampling research book (like Seymour Sudman's "Applied Survey Sampling") about the dangers of self-selected surveys. Virtually all web surveys fall into the same trapsite admins do it because they can not because it produces valid data.
nah. get an ICP vortex card. 256 meg cache using vanilla PC-133 ECC DRAM and you cna put upto 5 in one system with 5 channels each.
...by providing space for file caching.
I totally agree with what you're saying, actually, just pointing out that tons of RAM (which is very cheap these days) is sometimes a cheaper way to accomplish a similar end.
I couldn't find any 3ware controller cards. I found only this: 3Ware.
Bush's education improvements were