Slashdot Mirror
IDE, SCSI And Recording Everything
Raju writes: "For many years we were told that SCSI is superior to IDE. I always made my systems with SCSI and the others in the household got el-cheapo IDE disks. In the past SCSI beat IDE hands-down but now according to Simson Garfinkel, "today's IDE drives are significantly faster than SCSI drives". In the article at O'Reilly Network he talks about the tests they had run for storage of network data on disks. In the light of this article does anyone see any reason for going with SCSI in a desktop machine? For servers with heavy disk usage patterns it might be different due to command queuing." Disk types aren't what the article's really about, though -- it's a top-level look at network forensics (including advice on building a traffic-analysis system), and makes some interesting points about the unbalanced growth of storage and bandwidth.
It's about reliability and speed of SCSI.
Personally, I would stick to IDE, unless you plan on having a UPS on your system, and you don't plan on doing restarts. From what I have heard, SCSI isn't too fond of shutting down hard.. or maybe that is just a Sun problem...
"True knowledge exists in knowing that you know nothing"
I have two sets of IDE controllers on my system. Each disk I have has its own channel and controller. Because I get to use cheap IDE disks, the cost is much lower than SCSI and the performance is right on par with it. Its not the technology -- its how its applied and used in real life.
--Kevin
IDE may be faster than SCSI in some benchmark tests, but in multiple drive machines where IDE drives share controllers, SCSI will always be faster. Plus, access times and transfer speeds aren't everything. The fact that SCSI supports multiple IO commands at the same time is a major contribitor to the feel of speed.
The article's authors needed a way to store large amounts of network log data quickly - they're trying to capture packets in real time. For that kind of straightforward use (large volumes of data, only one user, no simultaneous read/writes) it's easy to see why IDE is more cost-effective and speedy, as the article states. However, when you add multiple users trying to write multiple drives simultaneously, the story changes, and the article simply doesn't address that.
What's your damage, Heather?
Plextor's SCSI CD-ROM drives are still the best out there for digital audio extraction.
Toms hardware has a review of Western Digital's new drive, the WD1200JB. 8 megs of cache . The article claims that the drive performs at par or better then Seagate's Barracuda ATA IV. IDE has come along way.
I want one!
Simson Garfinkel sounds like a cheap ripoff of Simon and Garfunkle.
according to Simson Garfinkel
God Bless you please, Mrs. Robinson, Jesus loves you more than you will know, whoa, whoa, whoa.
The advantage SCSI has over IDE is the multiple read, writes on the channel. So a single IDE drive compared to a Single SCSI drive may provide an advantage but you can put several SCSI devices on the same SCSI channel and make use of multiple read/writes. To obtain similar performance with IDE you would have to have a separate channel for every IDE drive. The biggest advantage to SCSI comes with raid. No sure you can get IDE Raid but compared to a good SCSI raid card, IDE raid sucks.
Like everything else about computing, it all in what you want to do with the machine. SCSI for mission-critical, IDE for everything else. Quite a simply formula; IDE's speed and throughput nowadays is fast/high enough to handle just about anything the usual user can throw at it.
...we are from the government - we are here to help...
As if the tens of thousands of times this has been hashed out weren't enough already...
The question of IDE vs. SCSI is not (or should not) be about speed. Really. There are nice, fast drives in each camp. If speed is all that matters to you, go with IDE, it'll be a lot cheaper.
So are there any advantages to SCSI? Sure. But not for the majority of people. SCSI's beauties are:
- You can hook a LOT of drives to one controller
- You can hook most any kind of device to the controller
- You can hook devices up both inside and outside of the case
- You can use much longer cables
- When the controller is waiting on one command, it can issue other commands while it's waiting
SCSI was designed for systems where you would either have many, many devices connected to the controller, or where many different processes (or users) would be accessing the hardware simultaneously - and in either of those situations, it *does* perform better than IDE. However, the portion of systems that will actually enter into that area are very, very few. In general, "if you have to ask, you don't need it."
As for straight speed, if you're looking for all-out throughput, don't rely on a single drive, get a RAID array - be it IDE or SCSI. By getting a faster drive, you can increase your throughput by what - 10%? 20%? A two-drive array will nearly double your throughput, and with quality controllers, it's fairly linear up through three to five drives - again, depending on the quality of the controller.
steve
Oh, you're not stuck, you're just unable to let go of the onion rings.
Didn't I read somewhere that Google uses IDE drives to host their database? Maybe that or it was archive.org. Companies with such huge investments in the technology have surely done their homework. If you take the cost, reliability, speed, and *availability* of both IDE and SCSI, IDE wins hands down no question.
-kwishot
SCSI is a very reliable system and is capable of handling high load from multiple requests. Never had the original CD-R writting issues that IDE had because the system was designed to handle constant through put. But like all things with price, IDE's affordability has won out over technical achievement and has slowly worked its way up, if you buy the right hardware. This is no different than from USB. Originally, it was pale compared to firewire. But, cheap gave it the edge to become a dominant standard. With USB 2.0, Firewire will be religated to niche before dying quietly. After all, for the price of a SCSI drive, I can easily buy three IDE's. It makes more sense to just mirror the data. However, where performance is the last word, SCSI will stay for now. At my work place, the amount of volume of data we process is staggering. There just isn't IDE hardware that handle the number of disks we need to have online.
My two cents,
Bel, the mostly sane.. "Of course I can't see anything! I'm standing on the shoulders of idiots." -- Me
i've been using strictly scsi in my systems since i built my first box. however recently i put together a system and simply due to price and the apparent gains in ide speed, i went with ata/100 7200 RPM ide drive instead of my usual scsi. i wish i could undo that decision, as the ide seems simply unresponsive at times. as soon as i can afford a nice scsi drive, i'm switching back.
-rp
I've noticed a HEAVY taxation imposed upon the CPU during IDE traffic, compared to SCSI.
Sure, transfer rates may be on par with SCSI, but when your IDE drive array starts pulling down huge data files while you are chunking your buddies in Quake3, your framerate is going to feel the pain.
I've not noticed this with SCSI, and it may be different under windows, which i do not run.
Also, SCSI is a more reliable.
Also, factor in the cost of additional controllers. 32 devices per SCSI controller. You want 32 IDE Devices, you will need 8 controllers. I think that sums up the cost difference right there.
I have one IDE HD, and 2 SCSI HD's in my desktop machine, and if I really crank on the IDE, you can see the slow down, where the SCSI HD's keep pace. I also have a CDRW, DVD, Scanner, Zip drive, and a CDROM on my SCSI chain - burning, ripping, and doing normal tasks does not slow my system down. Trying that on a system with IDE, tends to slow it down a bit. With my 29160 controller, I also have plenty of room for more HD's if I want.
I recently helped a friend install a CDRW (IDE) in his machine, and the first thing I tried was burning on the fly. Nope. Sorry. Need another IDE bus for that, so that the CDROM and CDRW are on seperate buses. Not true with SCSI. I burn on the fly all the time.
I guess I just like the overall flexibility with SCSI.
The IDE drive may be faster but the IDE BUS is slower when trying to read and write from multiple drives on the same IDE channel ..
I've never seen any reason for the use of a SCSI drive on any desktop machine. Its simply not cost effective and it doesnt provide any real, substantial benefit. On servers however, SCSI obviously has the upper hand when used in conjucntion with RAIDS etc... However, 3ware does offer some interesting IDE raid devices which have shown some intersting results. I've worked in the storage industry and generally there is little excitement about the 3ware technology. At least among the partners i've talked to who have used/tested it for deployment in their storage products.
F******* LOUDER! I CAN'T HEAR YOU! --Ozzy Osbourne
The main SCSI advantage is not that it's faster in I/O than IDE (although it used to be). The really big advantage was that (and I think still is), that on a server under heavy memory and processor load, SCSI will outperform IDE because most of the logic is moved off the CPU and onto the SCSI card. So when the CPU is pegged, IDE crawls, but SCSI keeps on chugging.
I think one of the big things is that processor speeds have kept on shooting up, meaning that while IDE has been considered a serious contender for small to mid- sized servers increasingly over the past few years, it's now becoming much more plausible to use it on higher scale systems.
If all you need is a fast single-user system, or a machine that performs a specialized task, IDE is fine and good. The drives are fast and huge for little money, and the caches are big enough to obfuscate some of the bottlenecks. TiVo uses IDE drives, even - as do most of the specialized NAS servers out there.
If you run a multi-user computer, high-end server, or a system where hardware reliability is at a premium, SCSI is still the way to go, though - but you pay a premium for it. Features like command queueing and disconnect/reconnect are really helpful when running a server that has to manage a heavy load, or a complex multi-user application. And the best RAID systems are still SCSI-based.
But if you are running a server box that runs some sort of brain-damaged inefficient server or client OS, IDE is more than enough for you.
-- Josh Turiel
"2. Do not eat iPod Shuffle."
It doesn't matter how fast they make hard drives. It still won't make Windows XP boot in a reasonable amount of time.
I am running a scsi desktop, I only do it to maxize the performence of the machine.
That's probably true. For example, you can buy a n 80GB western digital 7200RPM drive for $150. That is $1.88/GB. The only 7200RPM SCSI drive made these days is the Seagate Barracuda, which is $300 for 36GB: $8.33/GB.
That really isn't the point of SCSI though. I'll accept that IDE wins on a money-per-GB basis. But, IDE has a performance ceiling that SCSI doesn't have. You can't get 10000RPM and 15000RPM drives for IDE at any price, period.
There is a point, when building RAID systems, where SCSI exceeds IDE in the $-per-I/O-per-second metric. In desktop systems, you probably won't exceed this point. But if you intend to have stripe sets of 4 or more disks, SCSI will win the price wars again.
Anyway it really isn't a matter of SCSI being expensive and IDE being cheap. It's the drives that are expensive/cheap and it simply works out that expensive drives get SCSI connections and cheap drives get IDE connections.
P.S. Have fun trying to get you 4-disk IDE RAID all within 18 inches of your IDE controller :)
Apple has stopped using SCSI for their systems since the G3 IIRC and went for IDE instead. To compensate for external hot-pluggable drives, they've added Firewire.
So Apple is right IMHO. SCSI is obsolete. For high-speed bandwidth and external devices, use Firewire. For built-in HDs, use IDE.
There is nothing like the metal on metal sound of a high quality SCSI drive. Also you cant find an IDE drive to make the high pitch whine like the 10,000rpm Cheetah. The IDE drives make weak plastic sounds, or almost make no sound at all.
Shell out US$a few hundred, and you can have a super-speedy IDE RAID with mondo-capacity. And, if one of the disks breaks, ah, no big deal. Especially if it breaks 4 or 5 years from now: that disk that cost $250 in 1998 can be picked up for $50 today. Try finding an old SCSI drive that easily.
--
Repeal me, NOW!!!
Thank you.
For CD to CD copy you'd be hard pressed to beat the speed you could get with SCSI. The multiple IO at the same time makes that rock.
As x approaches total apathy I couldn't care less.
In high performance applications, SCSI is being used less in the real world. You see a lot of fibre channel, especially switched fabric architecture. With the way most sans will cache data, throughput is ridiculously fast in this architecture. As switched fabric becomes more and more affordable, you'll see it used more often in less demanding applications. now if only I didn't have to reboot to add disk to my Sun boxen. Damn sd driver, I guess I'll wait for Solaris 9.
Go away, or I will replace you with a very small shell script.
I have given multiple chances to IDE and IDE RAID to suffice for my wife's digital video editing system. It has consistently given me nothing but grief and headaches. Finally I gave up on it all and switched to SCSI.
However then my SCSI RAID card died and I had no $$$ to buy a new one. SCSI is extremely expensive.
So then I went to firewire.
And life has been smooth ever since.
mje0w!!!1!
SCSI hard drives have longer life expectancies than ATA drives.
For example, the Seagate Cheetah X15 36LP has MTBF of 1.2 million hours, whereas the Seagate Barracuda ATA IV has an MTBF of 0.6 million hours.
Longer life = better ROI
Phear The Phat Penguin
i think the only thing that matters now is how many drives you need.
ide is now faster, but has been limited to the amount of ide cards/motherboard you have...
granted, with the new abit max boards coming out, with 12 ide devices, that's not a problem...
if you need more than 12 hard drives, when you're building a perfectly NEW system, i would use SCSI... if not, just go with the 'new level' of motherboards coming out, and smack some IDE drives into the case....
now if i could only get a better power supply for all of them.
Runnin' On Empty
I recently read that one of Apples aims in developing firewire was to take-over the job of SCSI. Is this likely to happen ?
Will we one-day consolidate our internal and external buses ?
smak.
--
b0rk!
It would be great to see how much throughputdisks SCSI and IDE support. I worked at a place where they had a SCSI disk controller (a hub) where a lot of machines could connect to the array and see the disks as local disks. I have seen some cases where there is a fiber channel controller going into another fiber channel controller and the actual disks were simple scsi disks.
What metrics do people use? Price surely favors IDE, and proabably scalability -- it is very easy to install a dozen scsi controllers and connect to several hundreds of disks because of the several channels provided.
A quote here:
Connectivity: The ATA interface can only address two devices while SCSI can address eight devices (Narrow SCSI), 16 devices (Wide SCSI), 32 (Very Wide SCSI) or 126 (FireWire). There are also many peripherals available to SCSI only and not ATA.
The page I got this quote from is at http://www.acc.umu.se/~sagge/scsi_ide/
S
Lame stories are posted to Slashdot while Biker Gangs On Turf Warpath.
I wonder if John Ashcroft will call them as material witnesses or Donald Rumsfeld will
call in the Special Forces.
Thanks in advance.
Woot
IDE vs. SCSI is the big topic here when the authors are talking about how rapid advances in hardware vs. slow advances in bandwidth means it is becoming much more practical for more people to track everything that happens accross the internet.
The ramifications are important.
Also - how does this storage boon impact other kinds of surveillance?
This whole line of thought is a big part of making big brother a reality.
Just a thought.
.
It's hard to believe that's how Micronians are made. Why don't we see it right now by having you both kiss one another?
Just wait until Serial ATA devices start appearing on the market.. More bandwidth, round cables, and it's compatible with all current operating systems that support IDE.
RaGe
We're all just noise on the wires..
The author's analyzing huge volumes of network logs -- he needs a good analysis tool after capturing all that data on SCSI or IDE.
You should take a look at the addamark technologies LMS (log management system) product. It's a distributed, clustered database system that will load huge amounts of log data and run SQL/Perl queries against it much faster than regular commercial databases or other specialized tools. Check it out.
Yes, the delta between SCSI and EIDE drives performance seems to be shrinking, but I would take a 15k SCSI drive over a 7200RPM 8MB cache EIDE drive any day.
Just my $0.02
Lame stories are posted to Slashdot while Biker Gangs On Turf Warpath.
I wonder if John Ashcroft will call them as material witnesses or Donald Rumsfeld will
call in the Special Forces.
Thanks in advance.
Woot
The architecture of SCSI Harddrives or devices in any case, are designed to handle constant throughput and not need the CPU and IO bus to handle requests. SCSI is also a supieror way to use RAID. True, there's software RAIDs that appear to be a good solution, but they don't give you the same performance, or reliability.
On a side note, IDE's have high burst rates, SCSI is more sustained throughput. Also, my 5 year old SCSI drive still has a 4.5ms access time, most all high speed IDE drives don't have a access time like that, and when using RAID, that makes a large difference.
Please be precise and avoid using vague terms like "speed".
Ask yourself what you are trying to describe, is it the speed at which the drive spins, the rate of data transfer, etc.
This article is NOT a scsi vs ata comparison.
For the purposes of the author, some otherwise crucial performance metrics were not relevant. For intance, the inferior seek time of available ATA drives had minimal impact on his application, and thus were not weighed.
SCSI and ATA each have their place in different environments, and as long as the drive mfrs continue existing trends it is utterly useless to bicker about the relative merits of drive technology divorced from the context of specific applications.
From the look of the comments it would seem a slashdot editor actually read the story before posting it. That's a sign of the apocolypse if I've ever seen one!
I wish it were possible to moderate the initial article submission as being off-topic, because from what I have gathered from actually reading this excellent article is that the individual who submitted this story completely overlooked the primary topic on which the article was written...
The speed comparison of SCSI vs. IDE was most certainly referenced within the story context of the story; however, that was by no means the intended takeaway that the author had for his readers - it was but a supporting factoid of his other conclussions and thoughts. The article was a very written analysis, history and summation of the practice of Network Forensics. While it did cover a wide range of technologies (including hard disks) that aid in the collecting of such forensic intelligence, by no means was his observation of the increased speed of IDE drives intended to monopolize the reader's attention or be the central focus!!!
Even worse, the majority of posters have (unsurprisingly) focused on everything but the article's intended subject matter. Now ensues the typical flame-war of people supporting their preferred technology instead of having intelligent discourse concerning this exciting and evolving new field of I/T security...
Oh well...if you can't beat them, I suppose you might as well join them! For the record, my vote remains with the tried and true performance and quality of SCSI...
Beer is proof that God loves us and wants us to be happy. -- Benjamin Franklin
with IDE you have to kludge a lot and it's never reliable ( in my experience)
With SCSI external-device connection was _always_ given equal consideration in the protocol and feature designs.
With LVDifferential scsi you can get 25M !!!
I always wondered why the Beowulf people used simple network cards instead of doing a faster IP over SCSI layer for Linux. That would have been sweet, faster than gigabit at far lower cost ( we're talking 5 to 6 years ago, where gigabit was extremely expensive and SCSI was becoming very cheap.)
There is nothing quite like the auditory feedback of putting the swap file on a 'classic' drive, like a 300meg full height Western Digital; you really know when your machine is working.
I think you need a server operating system like FreeBSD that can handle server loads to see the difference. Single desktop non-scalable *linux systems probably aren't up to snuff.
according to Simson Garfinkel
Hello SCSI my old friend
It's getting very near the end
SpamNet - a spam blocker that really works
For the past five years I've run my system exclusively with SCSI components. When I first went out and bought a SCSI controller and a disk I paid a fortune for the privilege. At the time, the UW controller cost me $150 and a 9Gb IBM drive ran me another $300. The controller's SCSI BIOS added another 5 seconds to my boot time, and the IBM drive was full-height and loud as hell on account of it spinning at 10,000rpm. Regardless, I was a happy camper. I had consistently fast disk access, low latency and--best of all--I didn't get those annoying entire-system pauses while waiting for disk accesses to complete!
Over the years the benefit running SCSI decreased. First bus-mastering IDE channels came along and got rid of the annoying pauses. Then they started turning up the clock speeds with UDMA 66, 100, and so forth, until my aging SCSI drives could barely compete with even an average IDE drive.
Naturally, I did what any self-respecting bithead would do: I upgraded my SCSI components. By that time (circa 1999) the price gap between IDE and SCSI had narrowed somewhat (this was before IDE storage prices bottomed out) and I was able to purchase two 18Gb SCSI drives for a mere 25% than the equivalent IDE drives would cost me. And once again, I was happy with decent performance, low latency and high throughput.
Two weeks ago, I found myself scrabbling to free up a few megs and realized it was that time again, time to upgrade my storage. Looking at Pricewatch, I noticed that IDE drives are now cheaper than Big Macs and come in similarly absurdly-sized portions. Would you like 160Gb of space for your MP3s? No problem--they've got you covered, at $200 a pop! Meanwhile, relatively few vendors have stayed on the SCSI bandwagon, demand for SCSI drives is mostly limited to legacy systems that don't support an IDE bus, and a 160Gb SCSI drive will cost you $900.
In the face of this incredible price ratio, I did what any self-respecting bithead would do: I threw in the damn towel. Now I'm in a transitional period where I run 36Gb of fast UW SCSI storage and 160Gb of even faster IDE storage; I have a SCSI DVD-ROM drive, a SCSI CD burner, and an IDE DVD+RW burner, I/O controllers are fighting each other to the death to secure an interrupt, and the inside of my case looks like the aftermath of a tragic explosion at a cabling factory. I'm damned lucky my system is water-cooled, because I doubt any system fan could pull enough air through that morass of ribbon cables to make a difference in cooling.
The moral of the story: SCSI had its glory days, but it just ain't cost-effective anymore. And with Serial ATA looming on the horizon and promising God's own transfer rates, it just doesn't make any sense to buy SCSI.
"In the light of this article does anyone see any reason for going with SCSI in a desktop machine? "
:)
I do a lot of work with 3D Rendering and Digital Video, etc. I have tons of high quality footage that needs to be stored. The reason I'm running SCSI is because its' really easy to add new devices. SCSI has enough channels that you can have one card control a bunch of disks. I have 5 SCSI Drives at work and a couple of Firewire for transporting data around to other computers.
At home I have 1 SCSI and 2 IDE hard disks, and now an external Firewire drive. The SCSI drive is my performance capture drive. I have a 14 gigger that's reasonably fast, and an 80 gigger that's slow. The 80 gigger is for archival of the compressed video, or the uncompressed I don't need to get as quickly. Then I have the Firewire 80 gig drive (also slow) that I attach and do backups to occasionally. The drive stays off when it's not in use. I figure it's more reliable that way.
I can forsee the day coming before too long where I have only high performance IDE drives and Firewire drives, but no more SCSI.
"Derp de derp."
Is that they can have a lot more devices, and it isn't just limited to storage. MY personal system has 5 HD's, 1 cdrom, 1 zip, 1 scanner. All scsi. If it wasn't for that one thing, yea, ide would make scsi people stupid now days.
Not to mention, there is also the smartness of the scsi controller. I've used "good" usb scanners, and ittakes over you computer when scanner. With scsi, you can burn a cd, scan a picture, and play quake3 with out a hickup. Now, who would do all that? Well think enterprises. Think 14 15000 RPM scsi drives in a raid 5 (or what ever). Or think media people having to render imamges while saving to a file and other stuff.
Oh ya, and nothing like sending in an older (3 year old ) scsi drive for RMA, with no questions asks other than "how can I help you".
But I have to admit, these days I keep quesioning myself on my coninuation of buying scsi for home. The I just look at all the things I have in scsi, and think of how I would "try" to do it w/o it. And I can't.
Oh ya, somethign that you ide people can't do
1 10k rpm hd OS
1 10k rpm hd swap/tmp
1 10k rpm hd data
1 10k rpm hd applications/games
1 10k rpm hd mp3/downloads, etc
The spirit of resistance to government is so valuable on certain occasions that I wish it to be always kept alive
Its price versus Reliability for heavy usage. Speed is about the same and with raid 5 if one of the WD drives fail its no big deal. Its a Tyan MPX, Dual Athlon 1800+ MP's, 512 DDR, 2 Intel nics, tnt2 video(cheapest they had), 560watt gold connect Enermax PS ($180), promise sx6000 raid card. All for $2200 plus $150 for Mandrake 8.2. I'm sorry our new server Kicks ASS. Try doing that with SCSI for anywhere near that price. !!!!! 360 GIG !!!!!!
Need help finding the flow? http://www.myspace.com/naturalismandbalance
IDE RAID is a wonderful thing, to be sure. In cases where you are concerned about storage and reliability over random-access speed, Linux software RAID or an IDE RAID controller can do wonders.
We're putting together a pair of backup servers, and $90 80gb 5400rpm IDE drives in RAID-5 configuration make a fine solution. For the size of my company, 160gb or 240gb allows us to backup everything at a very attractive price.
The system I'm putting together is a rackmount PII with a Promise Ultra100TX2 (two-channel IDE) and 3 80gb Maxtor 5400RPM IDE drives. The drives run $90 at Pricewatch. My total cost was about $300 for 160gb of somewhat redundant storage. Each of these drives will do about 30MB/sec sequental access if you saturate the IDE bus 2 drives. 36MB/sec if you use 1 drive per channel. 3 drives almost max out the PCI bus in this machine.
And if you're just recording or backing up, even that speed is overkill.
-- Have you ever imagined a world with no hypothetical situations?
Although many people discuss the superiority of the SCSI protocol vs. the IDE protocol, this is not really the question.
Manufacturers produce the fastest disks on the planet on SCSI interfaces only. There are no 10K/15K RPM IDE discs, period. If one wants the lowest access time available today coupled with respectable transfer rates, one must purchase a 15K RPM drive, which are only available in SCSI interfaces.
For single-user access patterns, the author is correct to state that IDE drives have the lead today. StorageReview.com recently reviewed the latest 7200 RPM Seagate SCSI offering, and it was beaten down in single user tests by half a dozen of the newer IDE drives; however, when tested with server access patterns, it was the clear leader (excluding higher-RPM offerings, of course.) Still, 7200 RPM drives can't beat 15K RPM drives in any access pattern.
And I noticed the author was RAIDing drives -- 3ware's RAID products are very high quality, and their performance exceeds each and every other RAID card out there, SCSI or IDE interface. That surely contributed to his conclusion that current IDE drives are faster than their SCSI counterparts.
Join the NFSNET. Our prime goal is making little numbers out of big ones. http://www.nfsnet.org/
10-20%
Hell, with software raid alone and, say, 8 15k RPM U160 SCSI drives you can easily get a 100% speedup.
As soon as you go for hardware striping, your bus becomes the limiting factor, not the drives, that's why medium-high end boxen have differential SCSI i.e. 2 SCSI controllers speaking to 1 array.
Try 10 drives in a stripe over dual U160 SCSI controllers.
:-)
You will be impressed.
I'm out of my tree just now but please feel free to leave a banana.
For linear access to limited (less than 0.2 TB) storage I would always go with IDE due to the cost savings. For multitasking (multiple concurrent processes/users) with continual seeks over a large storage area, I would definately go with 15,000 RPM drives and RAID stripping using SCSI to achieve minial seek latency. The features needed to reduce latency definately fall in the SCSI court. Since this tends not to be a factor or adding additional RAM provides more performance per dollar, I would agree that IDE is almost always perferable for workstations when money is an important consideration.
</blatent plug>
Sam
I don't think this is the place for a USB 2.0 v. Firewire flame-fest. But they're two different standards for different things. I can't imagine that there won't be a 800MB/s or 1600MB/s implementation of Firewire--it's built into the spec.
USB 2.0 might make it easier for high-speed USB CD-RW drives or improve scanner data transfer. I don't think there's any way it can handle the volume of digital video that's the bread and butter of Firewire.
-AC
Ah yes, another IDE vs SCSI debate. That's my cue to bring up the same point I've made in the last 500 /. IDE vs. SCSI debates, and hope that this time someone actually has an answer.
Heavy use of SCSI drives does not noticably impact system performance. When I say "noticably," I mean those intermittent pauses a computer experiences during disk usage. That is, when you're moving your mouse and the pointer skids across the screen, making it incredibly difficult to get any work done. I absolutely hate this. If anyone knows of an IDE setup that will solve this problem, just THIS problem, I'll dump my ridiculously expensive Seagate X15 in a heartbeat. Until then, its worth it to me to shell out an extra $200/box and deal with smaller capacity drives.
Yes, I admit that I haven't used SCSI drives for the recent two years.
But the reason I preferred SCSI drives for some of the server systems was that SCSI drives were more reliable when multiple users were accessing the drive at once; the system with SCSI drives did not slow down so much as one with the IDE drives. For some reason, it seemed that SCSI controllers relieved some work from the CPU.
Whie the speed of the IDE drives didn't match SCSI ones that time, IDE was my choice when I built my personal system or small-sized server. IDE was simply much cheaper. However, when there is a possibility that multiple users use the server at the same time, I always chose for SCSI, because of its reliability.
Is it changed nowadays? Is IDE now faster and more *reliable* than SCSI? I heard from some of my friends that they always go for IDE when they build a computation server; they are ph.d. students from engineering schoool. I think they are right in the sense that multiple access to the drive is seldom an issue with scientific computing applications. But if they build one computational server for the entire lab people, I would definitely recommend SCSI to them, based on my previous experience. Now with the increased performance of IDE drives, do I need to switch to IDE drives for these cases? What are your opinions?
in the past SCSI beat IDE hands-down but now according to Simson Garfinkel...
OK! What do Simon and Garfunkel know about SCSI!
Maybe they should testify for Microsoft in the anti-trust case...
:)
....synchronous versus asynchronous io.
...'nuff said...
*==Sigma-Z==*
SCSI has parity checking, EIDE (UDMA) has CRC.
SCSI drives (from the same manufacturer) use exactly the same physical mechanism as EIDE ones, but with different controller cards (or sometimes, just different firmware and different physical connector)
SCSI and EIDE do exactly the same ECC mechanism and exactly the same reserved-bad-blocks mechanism.
FireWire Faq
Sure USB2.0 is about the same speed as FireWire, but FireWire hasn't been standing still - it's next version calls for speeds of 800Mbps and 1.2Gbps. There's even plans for fiber and wireless based versions.
However, even more import is that FireWire is PEER based. A computer is not required to transfer video from one device to another. There's already a bunch of video equipment that has FireWire support, camcorders as well as the Playstation 2(Sony calls it i.LINK instead of FireWire or IEEE 1394) come to mind.
While it might be possible to hack USB 2.0 for use without a computer, USB 2.0 wasn't designed for it. I suspect such a hack would be a successful as the "patched on security" we see in Windows.
Firewire vs USB 2.0 & SCSI vs IDE have something in common. Firewire and SCSI both have dedicated chipsets to handle devices, while IDE and USB 2.0 both rely almost soley on the processor to handle the brunt of IO commands and processing. That's why when I plug my firewire/ USB 2.0 portable drive into my firewire port it performs significantly faster than when it's plugged into the USB 2 port. Same drive. Sounds like USB 2 straight from intel's marketing department, you need a faster processor.
this would be an interesting discussion... i hope someone comments...
Amazing magic tricks
The Barracuda ATA IV is an IDE drive.
IDE still has a long way to go.
- A.P.
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
Yeah, if only those Beowulf dorks had
asked you. It would have rocked. You are
the man.
LOL.
Newsflash:
Some
LOL.
the article mentions 160 scsi harddisks ; there are scsi 320 hard disks now, running at a swift 15000rpm
ide will never ever be any better then scsi ; scsi is faster, has a higher fault tolerance, higher MTBF and it doesn't eat your cpu when you're copying a couple of gigs.
and that's that.
I have a firewire interface for IDE drives. Someone chime in on pro/con of 1394 vs. SCSI. I can tell you one thing for sure, with 120GB IDE drives costing a little over a buck a gig, you'd have to be insane to go SCSI for bulk personal storage. I can put a quarter terabyte of MP3's on my machine for under $300. Hell-yeah!
For many years we were told that SCSI is superior to IDE.
This is completely irrelevant unless you have an application which is tremendously hard drive bound, and you've done benchmarking to determine which type of drive or specific model of drive will work best for your purposes. Otherwise this is just the typical, meaningless fretting that some geeks have made a hobby of, such as buying a new, expensive video card so they can get 327fps in Quake instead of 270.
Maxtor doesn't make anything faster than 7200 RPM.
Looks like you've been screwed!
The latency of 10ms on ide and 5ms on scsi (typical) are by far the slowest things you have to deal with on a computer today (other than CD latencies). It hasent changed much in 5 or 10 years. You will notice a substantial speedup in a number of applications and windows functions.
This is kind of sad. Simson Garfinkel (co-auther of "Practical Unix & Internet Security") writes a nice article about collecting and analyzing network traffic. As an aside, he comments about SCSI versus IDE.
What ensues? A debate about SCSI vs IDE. Talk about your trolls.
Why aren't we discussing network forensics? Besides basic network security, there are a lot of ramifications to this as Garfinkel alludes to in the summary.
I have noticed this behavior on el cheapo IDE solutions. Get yourself a MB with a decent IDE controller (or buy a good IDE card).
Asus, Tyan, Intel. These are pretty decent MB solutions for relatively little money.
I am very small, utmostly microscopic.
There is a solution to your dilemma....
Heavy use of SCSI drives does not impact system performance because you have a SCSI controller handling most of the low-level IO.
I have an ASUS motherboard with an IDE RAID controller built onboard (Highpoint HPT370). I don't use the RAID aspect of it, but it's just as suitable as an IDE controller and allows me to use the maximum performance of my IDE hard drives while utilizing about 5% of my CPU. It is amazing how much more work I can get done while my hard drives are pegged.
All this for about $30 increased motherboard price...
From a forensics point of view, packets are more reliable than logs. The work that these products are doing is not even comparable, in terms of accuracy of network forensics. The packet based method gives you greater detail.
After a competent system compromise,the best you can hope for with distributed logfile analysis is discrepancy checking between systems and flagging of suspicious data. Having a packet history allows you to reconstruct the session, and figure out how what got recorded in your logs happened.
It is somewhat more difficult to go from logs back to packets, eh?
Never send anything unencrypted that you don't want to have appear in court.
I have a Cheetah X15 in my home computer and a Maxtor IDE in my work computer.
Sisoft Sandra has told me that my X15 is significantly slower than a 7200rpm IDE drive, but I can say from using them that my X15 (and even my previous 10000rpm Cheetah for that matter) absolutely runs circles around any IDE drive I have ever used.
I switched to SCSI in 1998 when I got my first Cheetah and I have never looked back. I would much rather have a fast hard drive than a huge drive that takes forever to access its data.
Plus, Cheetahs can warm the house on those cold winter days!
I have a woman and money. Life is good.
That's just great. I hope that anyone with half a brain who reads this article takes it with an incredibly large grain of salt. If you are using one or two drives, IDE might be comparable to SCSI (as long as your have the drives on separate channels!) for most "workstation" type applications. Any more than that, and SCSI is the way to go (or Fibre Channel...). Here are just a few reasons:
- tagged command queuing (multiple outstanding I/O requests to a single drive)
- disconnect (drive does not "hog the bus" while waiting for an I/O to complete)
- you can have up to 15 drives per channel (compared to 2 on IDE) with minimal performance impact
- 15,000 RPM SCSI drives are available, although they do require extensive cooling.
It really burns me when some idiot claims SCSI is dead just because he doesn't see any reason to use it on his POS desktop system. A friend of mine recently set up a PCI-X based system with 8 SCSI channels and lotsa drives, and benchmarked it at over a gigabyte a second transfer rates (yes, that's 1024+ MB/s). It'll be a long time before you see that with IDE anything.
(Serial-ATA does promise to bring many improvements to the low end of storage, but by the time it gets common, SCSI will be even further along with Ultra320, etc)
---- I made the Kessel Run in under 11 parsecs.
IDE has gotten a lot faster these days but there are still many flaws.
For one, most of the ATA133/ATA100 is a lot of hype. On long transphers (or any transphers that exceed the cache size of the drive) I have yet to see an IDE drive break 30-40 MB/s. In fact, testing an "ATA133" drive on an ATA133 controller vs. an ATA100 controller I saw no gain in speed. There was a gain from ATA66 because the ATA66 bus can't quite sustain 30-40MB/s constant.
Which brings me to another point, like all buses, the 66/100/133 is the peak allowed, it is usually not nearly that fast.
The drive speeds could be higher on IDE. You can get some top notch SCSI drives that run at 15,000 rpm. The best you find with IDE is 7200rpm. The drives would obviously be a little better at filling the bus if they had faster motors.
The IDE bus lacks any intelligence. It is the intelligence you are really paying for on SCSI. The command queues, multitasking bus, ect. ect.
Lastly, SCSI drives are obviously way more expensive, as are there controllers. Of course you are getting a higher quality (read=better built, not faster) product.
Basically what it comes down to in real world performance is no matter what you choise, IDE or SCSI, your disk drives will be the biggest bottleneck in your system by a long shot. If you run a single drive system, or have enough buses so you don't share them SCSI doesn't really provide enough to justify the cost on a desktop in my opinion.
About five years ago we noticed that IDE equipted PCs loaded linux (over our network) faster than SCSI equipted PCs. This was not the first time techs had told me this story as when the first IDE Macs came out about nine years ago the same thing was noticed.
The theory was that with one drive on the machine the SCSI bus had more overhead (the better to manage its greater abilities with) to deal with than the IDE drive.
Don't forget that scheduling of I/Os for different partitions on a drive plays a role in performance as well. It's often best to test your specific hardware/OS drivers/applications yourself and treat the popular press articles proclaiming one technology as best with a bit of skepticism
We've deployed hundreds of linux desktop machines without SCSI support saving $$$.
Can someone explain me why SCSI drives are more costly than IDE? I believe (I might be wrong) that many IDE and SCSI drives share the same mechanics and thus, its the electronics that change.
I can understand that 80's and 90's that SCSI electronics were expensive, but I would have expected that electronics prices would fall. How complex is a SCSI controller? Does it have a chip running at 600Mhz or something?!? (Guess not).
Any input about the reasons why SCSI $> IDE is welcomed.
Remember the year 2000? They promised us flying cars. They delivered the PT Cruiser...
but try doing interlocked operations on several ide drives. That's what RAID is about, and IDE will SUCK big time at it because you can't do interlocked io over several disks. With SCSI you can have operations going on on several devices on the same bus, but IDE is strickly a one at a time affair. So in a computer with a single disk IDE might win, but in a system needing a disk farm there is only one choice .... SCSI.
I'd like to see what benchmarks they are using to compare the performance of SCSI and IDE drives. For instance whether they are comparing sequential read/write or random read/write.
Take a comparison between Maxtors DiamondMax Plus D740x and a Fujitsu MAM3367MC/MP:
DiamondMax Plus D740x
7200 rpm
8.5ms average seek time
41MB/s(OD) to 25MB/s(ID) sequential transfer rates
145 IOMeter heavy load(should be about the same depth as the database index measurement)
Fujitsu MAM3367MC/MP:
15000 rpm
3.5ms average seek time
57MB/s(OD) to 44MB/s(ID) sequential transfer rates
364.34 IOMeter database index
Even if these specs are off by a few percent we can see that for sequential workloads with low queue depth that IDE and SCSI are about the same. IDE wins on price, SCSI wins on overall data integrity and drive reliability.
For random workloads however SCSI is far faster than IDE in general due to the much lower seek time. Random workloads also often tend to have very deep queue depths, sometimes 32+ commands are queued. IDE drives don't perform command queing. Because of a SCSI drives command queuing the drive can internally reorder seeks. This gives a very large performance gain. Rotational latency is a huge factor in seek time, this is why 15k rpm beats out 7200rpm with random workloads.
The article could have been a bit more descriptive when it described why they claim IDE drives perform better than SCSI although it appears clear that it is due to the typical sequential workload of their application.
Talk about scuzzy! What is it with all these old rock/folk has-beens making comebacks lately, anyway?
What? Simson Garfinkel? Who the hell is that? I thought it said... oh hell, never mind...
"Michael, I did nothing. I did absolutely nothing - and it was everything that I thought it could be."
In addition to the numerous advantages listed by previous posters, SCSI also includes the ability to "disconnect" a drive from the bus, through the software. In theory (or at least in my understanding), then the drive is effectively idle and you can unplug the device from the bus without damaging data on the drive, or corrupting data being sent to other drives.
Of course, those who would try this should be using another great feature of SCSI: the single connector attachment (SCA) plug, which allows SCSI drives to be hotswapped, and often assigned a SCSI ID on the fly.
While many have spoken about the ability for SCSI drives to be used in RAID configurations, a huge benefit is the fact that the drives can be swapped off of the bus/host without turning the host off. This is a huge boon for server environments, where uptime is king. IDE does not have any features like this.
SCSI also has the ability to be used in a "simple" cluster of two machines. Sorry, but I'm hardly an expert on this, so I can't fill in the specifics. But you basically have two identical machines each with a RAID controller, and then these are both hooked up to the same disk array. That way, if one machine goes down, the other still has the current file data.
New IDE drives have tagged command queueing also but Linux does not currently support it. From what I can remember at least IBM and WDC drives support it. Support was added around Linux 2.5.8 but was removed while it is being rewritten. It will probably reappear in a few weeks or maybe a little longer.
Per-cost, maybe. But it's silly to claim IDE is faster than SCSI with no qualifications (e.g. cheap SCSI).
Where I work, I have had to set up a few RAID systems, and the solution that we decided to work with was entirely based on IDE hard drives.
:
... 8 of them
... about ~$300)
Here is the solution I did recommend and implement (we are working with 120GB hard drives for now, but with the now available 160GB it can be a 1120 GB RAID)
1) The Hard Drives : ~$1200 (1TB solution for $1600)
MAXTOR 120GB about ~$150 a piece
2) 8 port Hardware RAID Controller: ~$900
3ware 7850 / ~$900
(This card is ATA133 compatible with 8 port Hot Swap/Hot spare independent ports and of course fully supported under Linux (kernel 2.4.18 is recommended))
(NB: we looked at the Promise solutions but preferred this solution over a SCSI Disk rack)
3) Case and Racks: $320
ENLIGHT 8950 / ~$160
(This case has 1x 3.5" floppy drive bay and 9x front 5.25" drive bays)
Removable HD Rack with Dual cooling Fans / 8x$20
(with ATA100 compatible cables)
4) Computer: old parts (new
Athlon 500 / 256MB / 20GB HD / NIC (no Monitor)
And works as we have about 4 of them with the solution presented here in different total HD size configurations.
-- Martial MICHEL
Beta is technically superior to VHS.
Novell Netware is technically superior to Windows NT.
SCSI is technically superior to IDE.
Does any of this matter to most of the market? Not really, since most people look primarily at up-front cost. I've been telling my customers (mainly small businesses) that mirrored IDE drives are the best value for general purpose data storage. The gap has narrowed; IDE definately makes more sense for most people (and even most servers) these days.
If I were specing out a system for high-end video editing, or a system that absoulutely had to process thousands of transactions a second, or a general purpose file or e-mail server that supported thousands of users, or a GIANT SAN, I'd go with SCSI. SCSI shines in really big storage pools, or in places where you absolutely need the fastest possible speed. But for most things, IDE undercuts SCSI by a longshot.
That said, there is one major problem with IDE, and it's not bandwidth (as most "higher-end" IDE-RAID controllers (such as some of the new ones by Adaptec) have multiple channels for multiple drives) - it's lack of VERY standard chipsets & APIs needed to access IDE block devices. The original spec has been hacked onto so many times that you're really at the mercy of the manufacturers' drivers for any "sophisticated" IDE implementations. This has gotten me into trouble several times. SCSI drivers tend to be more plentiful than high-end IDE drivers, and the testing cycles seem to be better because OS vendors actually care about them.
But again, people who buy IDE just on the technical merits of it may as well throw their money away. I wish the situation were different, but I don't think it will change unless drive vendors DRASTICALLY lower SCSI drive prices. Right now they're getting away with charging lots of extra dough simply because managers are hearing "SCSI is way better!" from their employees when purchasing hardware. That may have been true a few years ago, but it'll take a few years for the general consensus to swing in the other direction. (I really, really like SCSI too, and I think IDE sucks as a technology... but money talks) :(
Does anybody have any experience with SCSI adapters for IDE drives. The cost of an IDE drive + SCSI adapter seems lower than a normal SCSI drive.
The question's author specifically asks about desktop machines, and states that there are obvious server benefits to SCSI. It sounds like IDE is definitely the way to go for a single-user box.
You quitting proves that the karma kap worked. The most annoying of the whores shut up. --CmdrTaco
Just think about it, so many more billions of dollars are pouring into R&D for IDE drives for one simple reason, its a much larger market. Now I don't know if SCSI is out of the race yet, but you must accept that for the last couple years, for the first time ver, the race is actually close.
:).
SCSI is a limited use technology, its expensive and not many usual computers use it. Eventually, I think IDE will outpace it. Especially since motherboard manufactures now see the goodness in putting at least 4 channels of IDE on-board
Nope. That is HVD. LVD is less. 12m if you do not have SE devices on the bus. Even one SE device drops it further down to 1.5m. Check the FAQs on http://www.cablemakers.com. In btw: they are the only ones I found to supply HVD parts.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
I have a dual P3-850 (was a P2-450). Under heavy CPU load it remains suprisingly responsive. However, if it's under heavy disk load, it crawls, even though Ultra-ATA isn't very heavy on CPU utilisation.
My previous machine was a single PPro-200 with SCSI disks. Under heavy CPU load, it crawled horribly. However, under heavy disk load, it remained much more responsive than my current system.
Therefore I conclude that SCSI really does perform better, even if the drives themselves are matched on throughput and access times. I think most benchmarks suffer a little from tunnel-vision and focus only on the raw disk performance without really taking into consideration what it all means in real world situations.
I put up with the worse overall performance of IDE because it's so much cheaper. Of course, I'm up to my limit (4 devices) and need a new controller if I want to add anymore. And, I have to remember to be careful about tying up the IDE bus attached to my CD-RW when I'm burning discs. I can't see the last point being a problem with SCSI.
FreeBSD 4.3 flirted with turning off IDE write caching. This reduced write bandwidth to IDE disks but was considered necessary due to serious data consistency issues introduced by hard drive vendors. Basically the problem is that IDE drives lie about when a write completes. With IDE write caching turned on, IDE hard drives will not only write data to disk out of order, they will sometimes delay some of the blocks indefinitely when under heavy disk loads. A crash or power failure can result in serious filesystem corruption. So our default was changed to be safe. Unfortunately, the result was such a huge loss in performance that we caved in and changed the default back to on after the release.
[...]
There is a new experimental feature for IDE hard drives called hw.ata.tags (you also set this in the bootloader) which allows write caching to be safely turned on. This brings SCSI tagging features to IDE drives. As of this writing only IBM DPTA and DTLA drives support the feature. Warning! These drives apparently have quality control problems and I do not recommend purchasing them at this time.
So, SCSI is better both for performance and for data integrity.
IDE drives are fine in a desktop machine. It isn't likely to be heavily stressed and any reads and writes are likely to be from a single application at a time and a single user at a time with a CPU that is typically 99% idle. Such a user doesn't need the benefits of SCSI and the additional costs that the marketing people add.
If however you have 100 people all accessing different pieces of the disk, some reading some writing then IDE will just not cut the mustard. It requires too much CPU involvement. With SCSI the CPU just says here you handle this to the SCSI interface and gets on with something else instead. In addition, with SCSI I can have 15 devices on a single bus, with IDE, I can have 2.
So basically:
SCSI = scalability & heavy loads.
IDE = low cost & single user access.
Use the one appropriate to your application. For most people that'll be IDE, for other people chucking a lot of data around and lots of processes doing different things, SCSI would be better.
Just a quick rant about laptops. People think that a 1GHz laptop is as fast as a 1GHz desktop. It isn't. The laptop disks are designed with power management in mind and are often significantly slower than normal IDE even. So if your managment think that everyone should have laptops, tell them not to complain when their Oracle client runs like shit.
Government of the people, by corporate executives, for corporate profits.
What ever happened to the LINUX SCSI driver issues that came up about a year ago?
I was just wondering what happens if you RAID a bunch of RAID arrays. If you got a RAID SCSI controler from Adaptec or someone then hooked up a couple RAID controlers like one of these Accusys what happens?
I haven't seen any IDE controllers that sport a 64-bit/66 MHz PCI bus interface. SCSI already has PCI-X dual-channel U160/U320 controllers. Check out LSI Logic
IDE RAID is fine, it's cheap, but with newer IDE drives pushing 50 MB/sec (sustained) you could max out a standard PCI bus with three drives. Need more throughput? Then you're stuck waiting for PCI-X IDE RAID controllers, or at least 64-bit/66 MHz versions. And in the meantime, SCSI will just get faster.
Mainframe hard disks (historically SCSI) don't use remapped sectors. The drives are built to perfection. They are the top of the line. IDE drives are inferior, because the drives that have imperfections are sold off as IDE. This is because the bad sectors are remapped and hidden from the user.
Same old story people, OS/2 is the quality system but loses, Microsoft is the pile of junk yet sells to the masses. Likewise SCSI drives are the quality niche, IDE drives are the mass-marketed Microsoft-equivalent pile of junk with the bugs and flaws hidden.
While the manufacturing line at Seagate, IBM, Hitachi, Quantum, etc. take their SCSI drives *very* seriously, IDE is more like "yeah it's ok if we screw up a couple of sectors, couple of customers complain so what". The SCSI line failing is like Ford coming last at the Daytona, it's in a completely different league.
The reliability of SCSI drives outclasses that of IDE because the manufacturers discriminate during production (note IBM 120GXP fiasco did *not* affect its SCSI drives). If any cleanroom people can confirm my facts please reply.
A caveman dreams of being us, the incalculable power and riches. We dream of being Q, then what?
Apple still sells SCSI systems.
You order SCSI cards as a custom add-on.
I'm sure they sell plenty of SCSI cards in server and hi-end configurations for graphics editing. We have one for a server at my company. I also know someone who ordered one to better his DV editing experience.
Apple simply realized that not everyone who buys a desktop machine for word processing ( aka iMac ) needs SCSI, so why build it onto motherboards when IDE controllers/disks are so damn cheap?
How much cpu power does an IDE array need? Why don't they make a motherboard or raid card that has it's own CPU dedicated to IDE? An AMD 1800+ only costs like 110 and that is EXTREME overkill. I've never understood why IDE HAS to depend on the main CPU. You smarter types explain it to me :)
I've been a SCSI bigot since my Amiga days. Just 15 short years ago, all that was really available for consumer-level computers was SCSI, ESDI, and ST-506.
ST-506 was hardly an interface at all. You had to tell the BIOS the number of cylinders, heads, and sectors the drive had (sound familiar?), so that it could do the multiplication and convert logical block addresses into positioning information for the drive. You also had to enter the bad block list by hand, printed on a sticker affixed to the drive. An ST-506 interface was available for the Amiga-2000, and setting it up was predictably a bear.
SCSI saw its first consumer deployment on the Mac, and Amiga got it not too long after. No more CHS crap. No more typing in lists of bad blocks. All that intelligence was on the drive itself. Just plug the drive into the chain, tell the OS what SCSI address it had, and you were ready to start partitioning and using the drive.
So when it comes time for PCs to get intelligent drives, SCSI was the obvious choice. But no, they invent this new thing called IDE. What was different about it? As far as anyone could tell, the cable. You still had to feed CHS addresses at it; SCSI used LBA from the start. IDE drives from different manufacturers wouldn't work together; SCSI mandated interoperability. IDE now let you have two drives in your machine; SCSI already allowed up to seven.
IDE was touted as much cheaper, but it wasn't. SCSI and IDE drive prices were at near parity for years. Manufacturers were offering drives in both IDE and SCSI flavors (all other characteristics identical), with the SCSI flavor costing only ten dollars more (for a $600.00 drive, a typical price in those days, this was epsilon). It's only in the last few years or so that SCSI drive prices have skyrocketed for no readily discernable reason.
Add to that the fact that, even on a modern SCSI controller, all your old drives will still work. I have an old 600M 5-1/4-inch full-height Hewlett/Packard drive with a SCSI-I (asynchronous) interface. I plug it into the Adaptec AHA2940-U2W controller in my main rig, and Linux sees and mounts it just fine. Same with all my other old SCSI drives; I don't have to leave any of my data behind. It Just Works.
I also have an HP Omnibook 800CT laptop, which has SCSI built-in. All my drives work on that, too.
Apart from the artificially inflated costs, SCSI's only real headache is bus termination. But aside from that, the increased speed, flexibility, expandability, and reliability, for me, make SCSI an obvious choice.
Schwab
Editor, A1-AAA AmeriCaptions
a friggin joke. I use 68pin external seagate scsi drive which runs from a adaptec pci card.
my 2 udma100 maxtor na dibm drives run straight from my mb controller.
i have xp installed on my maxtor drive and an exact image of this drive backed up on the external scsi drive for transfering to my second computer which does not have an internal drive. the system boots HELLA faster from the external scsi drive than it does from the udma drives. FACT. PERIOD. have a nice day.
Ave Molech Setting
Under normal workstation or home PC use, IDE would be a cost effective method of connecting drives. You can buy a 100GB or so hard drive from Maxtor or WD for only a couple hundred dollars, while SCSI can cost more than double of what you can get IDE for. In addition, you must buy a SCSI controller to connect your drives to, this controller usually fits into one of your PCI expansion slots and can cost almost as much as a SCSI harddrive.
I have some recent first hand experience with this one. I built a 980GB RAID out of eight 160GB ATA drives on a 3ware card, and I was all ready to ship it, and but I forgot that we needed to copy 500GB of our data onto it first. That was early yesterday. It is still copying files as I type this. It's pretty bad when it takes several days to copy files onto an array at full bore network speeds, just to fill it halfway up. Never before has network speed/storage space been at such a high ratio before.
I've had enough abrasive sigs. Kittens are cute and fuzzy.
The fastest IDE hard drives (WD's 7200 RPM 100 & 120 gig drives) aren't faster than high end SCSI drives but they're nipping at the heels of some very respectable SCSI drives. When you consider the price difference, it takes a very stubborn (or ignorant) mindset to blindly say "only use SCSI for servers".
There are some situations where good SCSI drives still have a distinct advantage. That's generally in the arena of database servers and systems with a high I/O load on the drives. But if you're not going to be building a server for that kind of environment, you're doing yourself (and/or your company) a disservice by not at least considering using IDE drives.
A 240 gig RAID with 6 channel IDE controller, hot-swap bays, and a spare drive can be had for under $1500 using the fastest IDE drives available. Capacity can be doubled for another $600. That's about half a terabyte of usable storage with the ability to suffer the loss of a drive and rebuild itself on the fly for a little over $2k. Try and build that for under $4000 with SCSI. Doubling the capacity of the SCSI system will cost $2800 (assuming you started off with 73 gig drives). More than quadruple the cost of expanding the IDE RAID. To use a PHB phrase, "what's the ROI on that SCSI system?"
Unfortunately, many consultants are unwilling to even consider using IDE. "Gee. I've never built a server with IDE drives before. Sounds like it could work but...SCSI IS FOR SERVERS! LONG LIVE SCSI!"
All SCSI drives ship with write caching turned off. This is because they figure if you were willing to pay for a SCSI drive you must value your data.
:-)
All IDE drives ship with write caching turned on. This is because they want to win all the lame benchmarks that people run. Plus, they're mostly used in desktop PCs where you can blame disk corruption on lots of different things.
This difference alone can give a massive advantage to IDE on write-intensive applications, especially on those new WD drives that have 8MB of cache. Of course if the sytem powers off suddenly you could have a problem, but you have a UPS, right? Also, if you run IDE on Windows, just make sure you keep up with the Windows patches for cache problems:
Write Cache on IDE/ATAPI Disks Is Not Flushed on Shut Down (Q153296)
ScanDisk Runs Even Though Windows Shut Down Correctly (Q273017)
IMHO, the SCSI bus system is better than everything IDE/ATA can offer to date. It's not necessarily the devices that need to be put up against each other. Most recent SCSI disks in "acceptable" sizes are so expensive that you can easily build a RAID system from IDE disks for the same or even lower price. However what's really bad about IDE is the short bus. Face it, length and size do matter in some cases.
You can have a 12m LVD-SCSI bus with 15 devices plus controller running at full speed. But that's not desktop. You'll have trouble just cramming the disks in your average-sized tower, and you still need one or two additional PSUs to get them spinning. And now you take the sucker out for a LAN; but don't forget calling your chiropractor and get a reservation for the next two weeks straight.
Then there's IDE. With todays U-ATA133 specs you're limited to, like, 50cm bus length. Heck, that's about the height of a midi-tower! But it gets the job done. But no external devices for you, sorry. And you're down to 4 devices on your average motherboard, but most users can live with CD-ROM, CD-RW and one or two disks. With onboard RAID controllers coming up, there's an additional four disks possible and you can even plug in a separate DVD drive. You don't need a nuclear plant to get it running, you have lots of storage for a desktop machine and you can still carry it around. Perfect.
To sum it up, I think SCSI is still great, but it's losing on the desktop nowadays. The disks might last longer, it might be more flexible, but in the end, it's way too expensive and overkill. And then there's serial ATA on the horizon.
Fight hunger. Filet a politician and send him to a 3rd world country of your choice.
I/O is the key to everything in the SCSI vs IDE debate. IDE simply bogs DOWN under heavy I/O.
I ran a heavily-used mail relay. Initially, it used IDE. However, due to the load, the system was bogged down in I/O operations most of the time...system load wasn't high, but IDE was kicking the crap out of the machine.
Not so with SCSI. Even under heavy load, there was no I/O sluggishness.
Then there is the issue of reliability. I have YET to see an IDE drive with the reliability of a SCSI drive. SCSI drives cost more, and with good reason---they are made to last.
I have ultra 160 scsi on my desktop which I use for games and web development. The cost of the drives (3 IBM 10k 9.1 GB 80 MB per sec) was just over $100 on ebay. The Adaptec controller was almost $200 new on ebay.
"You do not support the root but the root supports you." - Romans 11:18
I did this test to see how badly my wd200bb drive was draggin down my raid. Not too bad, I think I'll keep it. Peace of mind is better than a few K/sec. :-)
0 ,3 13.4,1,16,2566,99,+++++,+++,+++++,+++,2628,99,++++ +,+++,5122,99
1 ,2 17.2,0,16,2632,99,+++++,+++,+++++,+++,2712,99,++++ +,+++,4757,99
Raid 1
Bonnie++ 1.02c
0 WD200bb
1 IBM 60GXP
496M,9813,69,26251,15,12396,11,12382,92,26699,1
Raid 1 w/ faulty
Bonnie++
0 WD200bb (faulty)
1 IBM 60GXP
496M,10562,74,27133,11,13380,9,12001,89,29111,1
3) SCSI allows you more drives - which means you have as much disk space as a slower IDE in 2 SCSI faster drives.
Which is easily aliviated by getting $20 IDE controllers. Hell, my last drive came with its own controler card (due to size)
I mean, just how many drives do you need?
autopr0n is like, down and stuff.
I guess you're right; when I sock my drive, it gives it quite a shock...
But wait, that's not all... included is a great review of RAID.
Make sure everyone's vote counts: Verified Voting
This should have been an article on drive noise.
And the vision, that was planted in my brain, still remained...
On the speed issue, I don't think this has been addressed... How many IDE I/O cards out there have slots for large amounts of RAM? How well do those caching SCSI controller cards work compared to an un-cached IDE? I dream of a controller card with 1GB of ram on it, and firmware that pre-loads the first 500MB of clusters into RAM (presumably system files and swap.)
Buy a pentium so you can reboot faster.
I was replying to part of the post by beldraen: So what is the point of your post?
information - what was the point of yours?
SCSI as a protocol is far superior (in terms of performance design, connectability, intelligence, and fault tolerance/scalability) than IDE (which essentially acts as a glorified signal converter). Regardless of what any benchmarks attempt "prove," SCSI does not present an overhead which inherently degrades single-user performance. Given the same drive mechanics and comparitive channel rates (ie. 80mb/sec - 160mb/sec LVD, or FC), SCSI disk performance will meet or exceed IDE disk performance, for any given single user application.
When you begin to involve more complex and real-world use of disk drives, the difference becomes tremendous. Think faster disks (15k rpm Seagates), switched fiber interconnects (running >200mb/sec), spindle synchronization, and intelligent command queueing. The added cost (which is usually insignificant compared to the cost of downtime, delayed I/Os, and maintenance) becomes a non-issue.
99% of the high performance computing industry chooses SCSI over IDE as their block device interface, time and time again, and there is a reason. To do so otherwise demonstrates a fundamental misunderstanding of storage interface technology.
A government is a body of people notably ungoverned - AC
Easy! Because corporate I.T. departments think nothing of shelling out huge coin for SCSI arrays. Let's face it, SCSI at home is nearly an impossibility in the face of gargantuan IDE drives for under a dollar a gigabyte. $400 will buy you about 18GB of SCSI storage if you include a controller. You can get about 180GB of IDE storage for the same price. Sure you can't hook up a SCSI scanner or SCSI CDROM to it, but why would you want to? Good USB scanners are everywhere, and IDE CDROMs lack nothing compared to their SCSI bretheren except an inflated cost.
So it has to be the corporations that are footing the SCSI bill and allowing the prices to remain inflated. The drive mechanisms and fabrication plants are IDENTICAL between IDE and SCSI. There is no special razzle-dazzle clean room with neato-keen testing equipment just for SCSI disk platters. It's the same! The only difference is in the electronics, and while SCSI chips cost more than IDE, we're talking the difference between a $10 IDE disk logic board and a $25 SCSI disk logic board, certainly not enough to account for the ridiculous pricing differences of SCSI vs. IDE.
Some have said that IDE is worse off because you can't get it in 10K and 15K RPM variants. To that I say "if you ask for it, they will build it". Your average consumer neither wants nor needs a banshee-screaming, egg-frying 10K or 15K drive. And as we've already seen, about the only thing RPM gets you is reduced rotational latency. It does NOT scale the transfer rate (i.e. a 15K drive's transfer rate is not double that of a 7200RPM drive) with RPM. Do we really care if the drive has a 0.5ms faster seek rate? On a database server, yes. On a desktop, or even a good workstation? No.
In the end they will lay their freedom at our feet and say to us, Make us your slaves, but feed us. - Fyodor Dostoyevsky
I have a dual p-III 1000Mhz system. I have a 7200RPM Seagate ultra2 18.1GB with 2MB cache and one 7200RPM Maxtor 40GB ata66 with 2MB cache. compiling a large project on the IDE takes 4 or 5 minutes total. Compiling the same project on the scsi drive takes 2 minutes.
That was all I needed to see to convince me that scsi was good for gcc.
SCSI has one really nice thing going for it...160MB/sec bandwidth. A bunch of drives in a raid 5 configuration can take advantage of Ultra 160's bandwidth, but it's a waste on the desktop since most desktop chipsets do not support 64-bit PCI at 66 MHZ. (Intel's i850 P4 chipset has a bug in it that limits the PCI bus to 90 MB/sec instead of the theoretical 133MB/sec.) If your platform doesn't have the bandwidth capability then there is no reason to spend the dough on lots of fast SCSI drives.
SCSI drives are (for the most part) built better and spin at higher speeds. 10k RPM drives can be had for $350. I have a 36GB 10k Seagate in my box right now. The SCSI drives that I buy come with a 5 yr warranty. You'll never see that kind of warranty on IDE drives. It has nothing to do with IDE vs. SCSI per se....it is all about margin. IDE drives have to be less than $200 to get anyone's attention in the retail space so they have lower build quality.
Most SCSI drives, on the other hand, are designed for higher-end applications where cost is less of a factor. Reliability and uptime are more important (read: servers). These drives cost more and therefore the manufacturer builds (and warranties) them better.
As far as single drive performance goes, the mechanicals of most SCSI drives and IDE drives are pretty similar. That is the limiting factor in a single drive installation...not the bus or the interface.
-ted
Anyone who believes "IDE hard drives" are faster than "SCSI hard drives" are out to lunch.
I'm fired. ;)
A government is a body of people notably ungoverned - AC
Funny you should mention that. I was playing Tribes2 on my linux box last night. Suddenly the game starts lagging real bad. Noone else is having a problem so I quite out of the game and run top. I also notice the HDD churning away. Lo and behold, I have an slocate process useing 85% cpu or more. Does remind me that I need to reschedule that cron jof for another time of day.
Just a Tuna in the Sea of Life
It is fairly rare for SCSI drives to have the same mechanism as IDE drives these days. The only place you find it is low end 7200 RPM SCSI drives, which are a disappearing breed (only Seagate is still introducing new 7200 RPM SCSI models). 10K and 15K RPM drives are on completely different mechanical platforms. In fact, even the media is different, as 10K drives typically have 3" platters and 15K drives 2.5" or 2.6" platters. (Easier to spin them that fast when the platters are smaller.)
Before 7200 RPM SCSI became the low end of the SCSI market, there were many 7200 RPM SCSI disks with different mechanisms than IDE disks. I once had some IBM UltraStar 9LPs; these had a 6.5ms seek time. Even today no IDE disk has a seek time that fast.
If you go back even further the story is much the same -- at any given point in time there are a few low end SCSI models which share mechanisms with high end IDE models, but never any mechanism sharing between high end SCSI and any form of IDE.
Also, SCSI drives typically have better soft and hard error rate specs, which indicates that they do not have exactly the same ECC.
Stop kidding yourselves. You will never give up your expensive trash anytime soon. The problem you are referring to has been dead for years now. You just like to dump an extra $200/box and have other people deal with you bringing it up all the time ("...did i tell you about my scsi box?").
Why can't manufactures make 15,000 rpm IDE drives?
I bet they would sell!
Power tends to corrupt, and absolute power corrupts absolutely.
you guys a missing another aspect, and those saying IDe is cheaper are quite incorrect in the long run...
SCSI drives from Seagate and IBM both have 5 year warranties... and most IDE drives have 1.. the SCSI drive will happily last 5 years without ever being turned off.. i'd like to see a IDE drive last like that...
i put all SCSI disks in my workstation simple because i know they're built better, will last longer and give me better performance without mucking around.
IDE drives are cheap and nasty and i could never trust data i wanted to them. Most ide drives these days seem to have trouble lasting 2 years. and the DOA rate of them is climbing every day.
~HCA~
I wasted alot of money 2 years ago on an Adaptec SCSI U2 card and some plextor drives for my home PC. Everyone told me this was the way to go to get good CD copies. Turns out that I could have been just fine with IDE and saved $300.
h tml
= Alle zA1Sport27&browselevel=roadbikes
Just say no to do SCSI for your home PC. Its like buying an expensive racing bike
http://www.litespeed.com/english/road/road.
when all you need is
http://www.specialized.com/SBCBkModel.jsp?san
MTTF, or Mean Time To Failure, is the time that a drive will fail after being installed.
MTBF, or Mean Time Between Failure, is MTTF plus the time to remove the drive, acquire another one, etc..
I guess what I'm saying is two things: good luck trying to find a SCSI drive in four or five years, and who will care by that time with the advancement of storage technology?
You can get 8MB buffers on the newest Western Digital drives. That gives a 7200RPM IDE drive nearly the raw performance of a 10000RPM SCSI disk. You can also run two in RAID0, which comes standard on many motherboards anymore nearly doubling the read/write performance again.
Unless you really need hotswappable disks or have huge I/O demands there isn't much point to SCSI on the desktop anymore imho.
Are there 15,000 RPM EIDE drives? I don't think so; the fastest I have seen is 7200 RPM and they are rare. What is the MTBF of the drive? EIDE may be faster at burning out. IBM gives a usage cycle of their drives that indicates they are designed for desktop use, not server use. This is usually reflected in the warrantee periods of the drives. 5 years is standard for SCSI, 2-3 for EIDE.
There's more !/$ on EIDE than SCSI, but the performance and reliability isn't there.
In any case, it's a matter of the appropriate use of technology.
Yet again, Simson shoots his mouth off without knowing the full story. Or maybe he does, and ignored it.
Hell, the article you link to even says as much.
Old and medieval windows are thicker towards the bottom because of the manufacturing processes involved; medieval-style windows made in the here-and-now using traditional methods are uneven right from the get-go.
Glass at room temperature is only *technically* a liquid because it doesn't crystallize. It doesn't really flow. In order for the glass in a window to sag enough to be visible, (assuming normal window usage conditions) the window would have to be older than the universe.
The article you linked to, by the way, questions heavily whether glass should be considered a liquid at all.
Just saw this today - EMC is now gearing
2 00 2.html
up to sell a betterfastercheaper ATA (aka IDE)
based disk frame - capable of up to a petabyte
http://www.nwfusion.com/news/2002/132134_04-29-
--chuck
Just ordered the parts yesterday. It'll be hot swappable (Promise SuperSwap) running NT server with 6 of those 7200 RPM, 120GB special edition caviars (8MB buffer per drive). I'm going to be using a Promise SX6000 (6 channels, one drive per channel) with 128MB memory in a RAID 5 configuration. Sure it slows down to do the parity, but the real bottleneck will be the dual port 10/100 NIC. After overhead, I expect a transfer rate over the wire of roughly 16MB/sec (200Mb - ~72Mb for overhead / 8). It'll be running a Athlon XP 2100+ with 512MB of DDR memory on a Abit KR7A Mobo. All the desktops in the company running WinXP Pro (About 20) will be backing up the entire contents of their drive to it nightly at midnight. I can't wait to run SETI and see if the performance dips when the backups are running. For those curious, case is a Enlight 8950 with 400W PSU, CDROM will connect to the regular IDE1, VGA is ATI AGP XPert 2000 (Inexpensive). I ordered the rackmount kit also.
Grand total was ~$2,800. Ironically, it'll be in the same rackmount as the Compaq ML530 which was purchased before I was hired. Each 36GB drive in that thing cost us $1,200. Times that by 18 or so. You don't even want to know how much they spent on the whole setup with Fibre connection to the server.
I talked to the senior technician at Promise and asked if running the drives in a master/slave configuration on their 2 channel and 4 channel cards would run slower than running one drive per channel when striping. I had always heard that with IDE, only one drive can be accessed at a time. He said that their controllers have a timing for each device so in theory it shouldn't matter, but in tinkering, he noticed a slight improvement when running with one device per channel.
I'd be interested to hear what others have experienceed with IDE RAID controllers versus SCSI RAID contrtollers when it comes down to transfer rate and performance. Do these controllers give the same effect of bogging down of the CPU like IDE desktops because of I/O? Or are they quite similar because of the IDE RAID controller with all the chips on the card?
I'm betting the packets were captured using NSM and then replayed using one of its associated applications . NSM logged packets to disk, and then at the end of the day, it went back through this data and created session summaries. Warning values based on string matches and a primitive anomolly detection engine flagged suspicious traffic. An analysis could then go back through the suspicious traffic and determine if the activity was truly malicious and more importantly, whether malicious activity was successful. From there an analyst could also determine what actions took place when unauthorized users had access to the system. All this without ever having to access the victim computer.
:)
Todd Heberlein created NSM and the application that "replayed" these sessions. He used to have Solaris binaries available on his site, but it seems to be offline as of late. Although the session-replay app (I cannot for the life of me remember the name) didn't provide a whole lot of analytical value, it was fun to play with. You could play at "normal speed" or speed it up. Telnet sessions (then the norm - ssh? what's ssh?) were fun to watch as typos would appear on the terminal and then dissappear as the "attacker" hit the backspace. Aaaah, them were the days
Bammkkkk
www.sguil.net
The Analyst Console for NSM
I think people generally get this wrong. THey think IDE is to SCSI as winmodems are to real modems, ie, all processing and hard work done by the cpu.
Not true.
The controller hardware in the case of IDE is on the drive itself. In scsi, there is actually less hardware and logic on the drive itself; those functions are moved to the controller.
The main bottleneck in IDE is the fact that it does not share a bus well. 2 drives on one channel is very hack-ish. SCSI, of course, was designed to be a bus, and to be efficient at what it does.
I'm confused.. you say you have 2 drives, striped, and then talk about copying big files between them? IF they are striped they are one volume, and you can't copy things between them.
I think the only reason IDE is more cpu intensive is because there are some functions, such as copying between drives, that scsi drives can do directly over the scsi bus with very little intervention from the cpu, and with ide, the cpu has to be involved in shuffling the data (or rather, with traditional ide controllers).
Oh yeah. A traditional ide 'controller' is not a controller at all. It's more along the lines of a raw i/o port. It has an i/o address, and a buffered (electrically speaking) set of i/o lines. It's even more 'basic' than your parallel port. I think you used to be able to build an IDE controller with a pair of 8255's.
Certainly they are a bit more complex now, but still, they are a raw i/o port, with no onboard functionality whatsoever.
The ide/scsi argument is still silly. If you want lots of drives working together, use scsi3 or fiberchannel. If you are at home, use IDE simply because it's way cheaper and certainly fast enough for you, especially with all the new funky ide controllers.
And it WAS faster for disk I/O back at the 233 MHz CPU times.
when I built my second PC I planned it to have a 233 MHz AMD CPU, 64 MB of RAM, CD burner and a scanner, all of this running linux. It was one of the best machines money could buy in Brasil at that time.
At that time SCSI made sense, because:
- IDE cd burners used to suck at that time
- SANE only had support to SCSI scanners
So I went shopping and came home with a Soyo motherboard, a K6 233 MHz, a Symbios SCSI card, an Umax scanner, a quantum 3.2 GB HD and an HP SCSI burner. Excelent machine.
Now we have nice things such as USB, Firewire, ultra fast CPUs and excelent IDE chipsets and linux supports all of them, so SCSI doesn't make sense for desktop anymore.
Now I have an IDE burner, IDE HD, USB scanner and USB printer, etc. and all works flawlessly.
SCSI nowadays is for SERVERS. where high the availability of RAID is a question of live and death, where reliable hotswap is neccessary and "details" such as extremelly high noise or subglacial cooling doesn't get into account because the machines will be locked in their own room.
now, if you REALLY want ultra-fast disks in your desktop... firewire is FASTER than SCSI. up to 400 MB/s.
What ? Me, worry ?
If you want your IDE hard drive to stop hogging your CPU, then turn on DMA for that drive if your controller and the drive supports it. If they do not, then get drives and controllers which do. If you do not do this, then do not bitch about your hd hogging up the CPU.
SCSI drives may be faster, but once you add other devices onto the SCSI chain, the performance decreases.
For 1/3 the price and a little creativity, you can build a very fast IDE raid array which will compete with the fastest SCSI drives out there.
Reliability? I have some reservations about a platter spinning at 15,000 RPM and how long it can continue to do so.
Firewire is 400Mbps, which is 50 MBps. That's faster than Ultra2 SCSI, but slower than Wide Ultra2, Ultra3 and Ultra160/320 SCSI. Check out this link for details. Firewire is still nice tech, and a fair bit smarter than USB2.0, but it's not the bandwidth king that SCSI is.
Ita erat quando hic adveni.
I have used SCSI devices for years in a broadcast enviroment. For the most part, I've had really good reliability, but in one case, I junked the whole SCSI array in favor of IDE. The station I work for uses an ADTEC MPEG 2 playback system for commercial announcments. It contained 6 SCSI TWO Quantum HDs in an external RACK mount unit with redundant PSs. After about 6 months of operation, drives started failing at the rate of one per month. This gets very old when your paycheck depends on spots running correctly! I opened the playback unit one day and discovered that it contained 4 IDE controllers on the MB. I promptly purchased 2 30gb MAXTOR drives and replace the SCSI array. That was 18 months ago and I've not had a single failure or commercial going 'blocky' since. Perhaps now would be a great time to knock on wood! ;)
I don't buy IDE drives for my servers because I require features that simply don't exist in the IDE world - RAID5 using more than four disks, and hot-swap.
Performance means next to nothing on the day you have a hard disk failure. On that (rather dark, unfortunate) day, all you really want is to be able to remove the failed disk without an outage, and plug in a fresh one. Show me an IDE RAID solution that supports these requirements, and I'll change my mind.
As for overall speed, there's no doubt in my mind that IDE and SCSI are pretty much equal in a single disk configuration. If you take cost into account, SCSI is a total waste. Going to multiple disks is still hard to justify, unless you bring RAID into the equation.
As with all computer system purchases, the requirements almost always define the solution. Define your requirements carefully, with the right priorities in mind, and the answer will almost always be obvious.
Incidently, I buy SCSI based RAID solutions for production servers, single IDE disks for development (non-production, non-backup servers), and IDE for all workstations (including some extremely high-end graphical workstations).
--
This would be my 2 cents worth, except we don't have 2 cent pieces in Australia any more, and rounding it down makes my opinion worthless!
I totally agree, the only problem at the moment is still cost (mainly because I don't have money for a cup of coffee, so even IDE are expensive for me).
I think i am going to perhaps even go with RAID-5 on my system (as soon as I put another processor in). I have a Raidport on my supermicro p6dgu and I am going to order the Adaptec raid controller for it. It has 64mb of hardware cache then. I will just throw 8 drives or something in it and it will be soo much better than my 4 crappy ide drives in there now.
SCSI is also SOOO much better for CDR drives. Ever try to do a bit-for-bit copy with an IDE drive, don't think so! Why did plextor stop at 12x for SCSI!?!?!?
Tibbon
A dual channel 160UW system with two 10K (or faster!) drives will blow away any dual drive IDE system. But it would be more expensive.
SCSI has multitasking capability (you can give multiple commands to each of multiple devices and have them processed as each device gets time). Add a CD-Writer or even CD-ROM to an IDE system and watch what happens if you try to access both the disk and CD. The Multimedia command set was originally SCSI (which is one of the reasons there is an IDE-SCSI option/module in the linux kernel).
You can also put 15 devices per wide SCSI bus, but still can only do two per IDE.
IDE is simpler and thus cheaper, but has a limited performance envelope, i.e. degrades under a number of conditions common to certain environments like servers. It isn't really a VHS v.s. Beta. CRT v.s. LCD might be closer. Or Sedan v.s. Pickup truck - If your job involves taking clients to lunch v.s. going to Lowes or Home Depot to get stuff.
Ever used Protools? The new version records 64 tracks at 192khz @ 24bit depth. Let me do my math here, 192khz (times per second) * 24 bit * 64 track == 194,912,000 bits/sec, I think that would overrun most of your IDE drives!!! Yes, digidesign (www.digidesign.com) does require SCSI for anything above 24 tracks at 44.1khz at 24 bit!
/. uses 9gb SCSI drives on most of their servers?
Ever needed faster access to things in 3d studio max? What about Final Cut Pro? Can you tell me that full video works great on IDE? Don't think so? Ever had a server with a few hundred (or better thousand) acessing it? Ever notice that
USB 2.0/Firewire could be interesting however if they started to put decent drives in the units (seagate i swear by!!!)
Tibbon
Umm, most supermicro boards support the 64bit pci slots. Those have to be the best mobo's around. I love mine.
Troll this fucker is. He is a fucking ass. He loves YASSER FUCK TOWELFAT. FUCK YOU PUSSY FUCK!
Arafat tours his battered realm, but there is no hero's welcome from his people
By Phil Reeves in Ramallah, West Bank
03 May 2002
Yasser Arafat had the air of an ageing thespian stepping out of the wings to savour a curtain call and drink in the warmth of his fans before taking a triumphant bow. But, as he strutted anew on the world stage, the applause was patchy and so were the crowds.
His aides and bodyguards thronged loyally round him, no less excited than they had been the night before, when Israel withdrew its tanks and snipers to end the month-long siege of his crumbling and bullet-spattered compound, allowing them to leave.
The press engulfed him, lapping up his every word and chasing his convoy of Jeeps and Mercedes, still caked in a month's worth of dust. He flashed V-for-victory signs and led prayers before a makeshift grave at the hospital - the only place officials could find to bury 15 people who were killed in the first few days after the Israeli army's arrival, when the town was under curfew.
He inspected the Ministry of Education, peering indignantly at the computers smashed by Israeli soldiers, and held forth about the "crimes against religion" committed by Israel in Bethlehem's Church of the Nativity - where another Palestinian was shot dead yesterday. He talked, as ever, of how he wanted to return to "the peace of the brave" that he sought to create with his "partner" Yitzhak Rabin.
But as the Palestinian leader - the "president" of an unborn and increasingly unlikely state - travelled around Ramallah with his entourage of Palestinian Authority officials, he received no hero's welcome. There was ambivalence and weariness, uncertainty - and even resentment.
Worn down by weeks of living under an Israeli military curfew, and depressed by the scale of the destruction the Israeli army left behind, this West Bank town of 40,000 did not cheer much at Mr Arafat's sudden appearance. His popularity, which blossomed during his months of confinement, seemed to have faded, blighted by the public's view that he has made a murky and dishonourable deal in return for his release.
Palestinians on the streets yesterday were subdued. They have two principal objections to the agreement, brokered by the United States and Britain. There is also deep distrust that any deal made in the current atmosphere of ill will, especially with Ariel Sharon, would hold for long.
On the second point, they were soon proved right: Mr Arafat received an explicit guarantee from the US and Britain that, if he met his side of the bargain, he would be allowed to travel abroad. But the ink was barely dry before Mr Sharon threatened to prevent Mr Arafat's return if he went overseas. And - though the Israeli army pulled back from Ramallah - it still blockaded the town from its edge, just as Israel is doing across the West Bank.
But, above all, Palestinians complained that Mr Arafat had betrayed them by handing over six Palestinians to be held in a prison in Jericho, where their Palestinians guards will be under the constant supervision of British and American wardens.
Four of the men are fêted by the Palestinians as members of an assassination squad from the Popular Front for the Liberation of Palestine (PFLP) that killed an Israeli minister widely loathed among Arabs for his racist views and whose death followed Israel's assassination of the PFLP's top man. The other two are the leader of the PFLP, Ahmed Saadat, and Fuad Shobaki, a senior Palestinian Authority official accused of bankrolling arms smuggling. All six are widely seen by Palestinians as heroes of the intifada, leading a legitimate armed struggle against occupation.
"The Palestinian people are all against this," said Loa al-Siyour, serving shwarma sandwiches in one of Ramallah's many cafés, thriving again now. "These people struggled for our freedom and then they put them in jail."
A few doors along, Mohammed - he would not give his full name - demanded to know why he should feel happy. "We are not even back to the point we were before the Israelis invaded," he said. "We are still living under closure. And now we have six of our people behind bars, being looked after by the Americans, Israel's friends."
There was unhappiness, too, over the cancellation by the UN secretary general, Kofi Annan, of a UN fact-finding mission to look into the Israeli atrocities in the Jenin refugee camp. UN, American and British officials insisted that the two issues - the mission and the terms of Mr Arafat's release - were unconnected.
But this has not convinced Palestinians, and nor is it likely to. Asked about this by The Independent, Mr Arafat reverted to type: "Jeningrad?" he asked. Stalingrad, one was tempted to say, was on a somewhat more terrible scale.
FUCK YOU TOWEL LOVER. You communist pig.
Here's the really stupid thing about all this: Although SCSI is hands down a technically superior interface it is not THE standard simply because some marketing bozos thought they should keep IDE along for the ride to better price differentiate the marketplace. I, for one, do not believe that SCSI electronics are much (if any) more expensive to produce than IDE--especially now that modern IDE RAID controllers are getting more sophisticated. There's nothing special about etching silicon for SCSI controllers vs. IDE, that should demand such a disgustingly huge price difference. In fact, the SCSI drives themselves ought to be less expensive than IDE since most of the interface logic is on the host adapter.
If only engineers ruled this world. (-;
"today's IDE drives are significantly faster than SCSI drives".
Fastest ATA drive at the moment transfers up to almost 50MB/s near the begining of the disk, fastest SCSI drive transfers at 60MB/s near the begining.
The SCSI drive has a measured seek time of 3.9mS whereas the ATA drive has only 8.9mS.
Factor in SCSI abilities to be able to queue commands, automatically re-order them and use the bus for more than one thing at a time and you'll see that SCSI is still the fastest. Especially so for IO intensive applications where lots of small transfers are requested.
Whether the price difference is worth it or not for your application is what really matters.
The Western Digital Caviar WD1200JB is one awesome ATA drive though.
War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
FYI, the disks are practically same. Some manufacturers may substitute disks from the best production line for SCSI-drives and some may not. Some may give 5-year warranty for SCSI, or just 3-year for both SCSI & IDE. The disks are equipped with either IDE or SCSI chipsets (in SCSI's case, that's less logic than with IDE), so that's not much of a difference.
It's the bus. IDE stalls your machine if you're doing heavy disk operations on any system with one or more IDE disks. A copying of a large file will do. Switch on DMA and 32-bit controls with hdparm, little help. Reading a cd or copying large files or just having constant disk activity means that you'll be waiting for IDE queries most of the time before the queue is getting shorter. With SCSI, you probably can't think of such usage on a desktop computer that would halt an SCSI bus. I've got experience on both, and with SCSI, the speed of the data storage stems much more from the speed of the disk itself -- with IDE, you'll always have the same bus trouble with heavy disk usage no matter what UDMA/xxx chipsets you have. And before you say, an example of a more frequent disk usage occurs everytime you boot up your computer and programs are being started and drivers being read at the same time. SCSI machines generally boot up a little faster than IDE machines, according to my experiences.
Does having two devices on the same IDE bus slow down transfer rates?
Thanks
Nevertheless, it does seem to be the ugly truth, at least for straightforward read/write tests in a single-user environment.
You essentially have one process doing sequential writes to disk. As almost everybody who has commented intelligently on the difference between SCSI and IDE seems to agree, this is the kind of situation that's going to favor IDE. I'd go as far as to say that it's probably close to IDE's dream application.
No big surprise here, for me.
ON the other hand, 25,000 users trying to POP their mail will probably beat those same IDE drives into submission faster than you can say "thrash the system".
Sometimes boldness is in fashion. Sometimes only the brave will be bold.
For performance the short, short summary is that IDE wins in single threading IO due to small, low latency command overhead while SCSI wins out for multi threaded IO due to command queueing. Sure, it is in the IDE specs too these days but it is rarely implemented.
Check out the HOWTO and if you see a mistake do contact the autor, the LDT relies on feedback from you the readers and experts.
for fark sake... ide sucks and SCSI rocks.. wake up nerds and unite against IDE.
The only reason that I have a SCSI interface in my PC is to hook up my sampler to my PC to transfer sampledata from the sampleeditor software to the sampler. Face it it' is handier to edit sounds on a computer screen rather than the tiny lcd screen of the sampler. It's also very handy to share my ZIP and CDROM drive between the PC and the sampler. I think that's the only but very valid reason to use SCSI. Can you imgaine to transfer sample date over a MIDI lead ... ha.
IDE supports that too, at least some drives: IBM deskstars and WD Expert series. Linux 2.5 and FreeBSD supports queueing on these drives.
When I've bought cutting-edge SCSI equipment, I felt gouged. When I've bought previous generation SCSI equipment, I felt like I was getting what I paid for.
I think that the cutting-edge SCSI drives prices are used to offset the R&D costs. By the time IDE drives reach the same point, the research is already paid for, manufacturing details are already solved, and the company is ready for traditional mass-marketing.
Another factor is reliability. I don't believe that SCSI and IDE drives have the same mechanisms, unless you look at old or crappy SCSI drives and cutting-edge IDE drives. Regardless, you get a longer warrenty with SCSI drives than with IDE drives, and you're probably paying something for that. If the drive mechanisms are indeed the same, I expect that the best mechanisms from a lot end up in SCSI drives, and the rest end up in IDE drives. Just as microproccessor speeds are determined by testing.
-Paul Komarek
Or, to put it another way, it's just much easier to make a decent living off a 100 cheap IDE drives than the rare SCSI.
I'm in a Unix state of mind.
Hello all, I am hope the vast wisdom of all of us can slove this problem I am having, I just bought this used system. Good system, however it has a scsi drive in it. AtlasV - incase your wondering, anywho this is my first SCSI drive ever, I've never messed with them. I currenty have a 46gig IDE drive - thats were all my data is, I am unable to burn CD's mind you. Here is what I am triing to do, I am triing to install the IDE drive and use it as a slave - transfer data, I am not having any luck however. When I install the ide drive it takes over and you can't see the SCSI drive. Likewise if I change jumper setting with the SCSI drive, also when I change the Bios. Thats allows the SCSI drive to boot and than I cant see the IDE drive. I think you see where I am going with this. Any Help or can this even be done. I really am starting to think it cant be done. I have sreached and sreach google, message borad etc, maybe I am just looking in the wrong places. Any help would Be Just Peachy. Not to mention taken with the WOOOHOOOO that would follow. Thanks all!@! P.s. i know this sounds like "why the hell does this person want to do this, answer = because I don't know if it can be done, learning thing!@!
Taste_Death_Live_Life = read it Live By IT!@!
...which is why I never have considered SCSI. With disk sizes growing (my 45gb disk was huge, now my 100gb is huge), I can increase capacity all the time (=better capacity), to higher data densities (=faster), more cache (=faster). SCSI would be better at purchase, hands down. But then I'd have to keep the same disk so long, it'd be inferior to the newer IDE drives I would have bought if I'd gone with IDE. I'd rather take a 200$ drive each year than a 600$ drive every three years, it's that simple.
Kjella
Live today, because you never know what tomorrow brings
I was told several years ago that the SCSI?IDE diffefence was mainly used by manufacturers rather as a "professional/consumer" flag.
The silicon cost (negligible nowadaya) and raw I/O capability of the two interfaces are effectively the same. If you have the same media rotating at the same speed, you get the same raw throughput. The article described essentially bulk logging application, probably driven buy a single thread (i.e. no room for overlapped IO or seek optimisation). The advantages of Scsi won't show up in this case, so (MTBF) aside, you might as well use IDE.
Typically, Scsi drives have faster seeks and larger cache rams. The Scsi interface (given a good driver, which is not automatic) allows much more overlapping of transfers. The build standard is usually higher, so SCSI drives have much higer MTBF (1,250,000 hrs vs 250,000 hrs last time I looked). 15,000rpm drives are only available (AFAIK) in Scsi.
For most applications, throughput is overwhelmingly dominated by seek time and rotational latency. Scsi will win by miles here, bith because of non-scsi-related features for which Scsi is a flag (haster seeks) and scsi-related features such as much more ovelapping.
(In on application, in which I had already sorted the transfers inot order, I still found a 25% increas in throughput by giving the Scsi drive a whole batch of commands to get on with in parallel).
But the head and oxide technology are the same, so for bulk linear logging, IDE will match the performance of Scsi, and beat the price.
Horses for courses.
Consciousness is an illusion caused by an excess of self consciousness.
You need a handle on the distribution--which is probably approximately normal from the Central Limit Theorem--and other parameters. Assuming nrormality, MTBF and variance (or standard deviation) will let you use a chi-squared distribution with n degrees of freedom (where n is the number of drives) to come up with mtbf for the group (again, defining failure as "at least one").
THe statistics get uglier when failure is "at least two":)
hawk
"Straightforward read/write tests in a single user environment." hardly specifies whether the tests are sequential or random. Single user could mean a single process or dozens of processes on the same machine. Single user only really means that the machine has a chance of not being loaded 100% at any given time, something that has to be accounted for in a system used by many multiple users.
But from everything I've been able to gather, the IDE implementation of TCQ is Broken As Designed compared to SCSI. In a SCSI system, the drive can process commands and then notify the SCSI controller that a command has been completed.
On an IDE system, however, the IDE controller has to poll the disk periodically to see if any commands have been completed. The drive has no way to notify the controller that data is ready and waiting.
It's the difference between a polled and interrupt-driven system. Polling can be fast, if it's very carefully done, but interrupt-driven systems are easy to make fast.
Don't get me wrong, it's a nice improvement to IDE, and it does narrow the gap somewhat, but as its always been, for high-end multitasking stuff SCSI is still the champ.
PHEM - party like it's 1997-2003!
I hunted all over the net looking for benchmarks of SCSI systems versus IDE systems, and couldn't find a damn thing. Sure, people benchmark SCSI hard drives versus IDE hard drives, but nobody ever bothers to benchmark SCSI RAID versus IDE RAID.
I got so sick of it I said what the hell, and ordered up a pair of raid arrays for doing my own tests.
Test system configuration:
Supermicro P4DP6 Mainboard with two Intel Xeon 2.2GHz processors, and 4GB of memory.
Microsoft Windows 2000 Server
No I did not have time to test this under Linux before I had to get these bad boys ready for prouction. I doubt the benchmark results would have been much different, I've seen 3Ware running under Linux at Linuxworld and they kick some serious ass.
IDE setup:
3Ware Escalade 7450 Raid Controller
Three Seagate Barracuda ATA IV Drives (7200RPM 9.5ms access time)
This was set up in a RAID5 configuration.
SCSI Setup:
Mylex AcceleRAID 352 Raid Controller
Three Seagate Cheetah LVD160 Drives (15,000RPM 3.6ms access time)
This was also set up in a RAID5 configuration.
Benchmarking utility used:
HD Tach 2.52
Here's the bottom line I got out of my benchmark tests.
SCSI Performance
CPU Usage: 2.1%, Access time 6.1ms
Read Speed: Max 37.6MB/s Min 11.6MB/s Avg 30.8MB/s
Write Speed: Max 8.5MB/S Min 5.4 MB/s Avg 7.5 MB/s
IDE Performance
CPU Usage: 3.1%, Access time 14.2ms
Read Speed: Max 31.8MB/s Min 4.3 MB/S Avg 18.3MB/s
Write Speed: Max 48.7MB/s Min 12.3MB/s Avg 36.4MB/s
I was a bit shcoked to see the IDE do so well. It annihilated the SCSI in terms of sheer write performance but lagged behind a bit from the read performance. CPU use isn't a factor for most people, who really cares if you lose 1% more of your CPU to the IDE compared to the SCSI.
Those 15,000 RPM drives were loud as jet engines, and they got hot enough that I was thinking of cooking some bacon strips on them. They were too hot to touch. The IDE drives on the other hand were barely audible even with the case off, and remained completely cool to the touch through the whole test without even a fan on them. You tell me which of those two types of drives is going to have a longer MTBF...
I didn't even use high performance IDE drives for that test. I'd also like to point out that the Mylex card was 66MHz/64bit, whereas the 3Ware card was 33MHz/64bit... so the 3Ware card was holding its own even though it was running at a slower rate of speed. I wonder what will happen in future generations of these controllers when they turn up the speed and improve the code...
Cost... I coul have built three of those IDE Raid systems for the price of that one SCSI system.
Space... The IDE drives were 80GB, the SCSI were 36GB. IDE owns SCSI in terms of space. We have some bigass databases where I work so that's actually fairly important to us.
Unless you REALLY need that 6.1ms access time or the extra ~20% in read performance you are far, far better off with an IDE Raid at this point.
The guys at Toms and Anandtech really need to do a major article on this stuff...
For the skeptical, here's a link to the screenshots of the HDTach benchmarks I ran. Be GENTLE guys we do not have tons of bandwidth for this...
IDE vs SCSI
IDE is on the top, SCSI is on the bottom. Interesting how SCSI is fairly linear but IDE is really sloppy and just running all over the place.
Hell is being intelligent in a world full of idiots.
Most IDE drives have 9-10ms seek times; whereas
scsi is generally around 5. That alone is worth
a lot in real world performance...
Simon & Garfunkel? What do they know about IDE and SCSI?
Although single user systems can have multiple processes at any given time, many/most single user apps usually have one process (max) accessing the disk at a time. A UNIX box being used as a dedicated network log would be such an application.
Granted -- a one-line summary doesn't include the source code, but it does indicate the kind of testing methadology they're likely to be using.
(btw: I consider my own box single user (me) even though it has 100% utilization -- I run Seti-at-home. The disk utilization is still single-user in nature (seti rarely accesses the disk)
Sometimes boldness is in fashion. Sometimes only the brave will be bold.
The fastest raid controller on the planet is IDE. (this might have changed, but I know it was the fastest not too long ago)
http://www.3ware.com/products/Escalade7000DS4-1
Real measured sustained 180MB/s reads and 127MB/s writes.
Of course that is a pricy card.
==>Lazn
Write back caching on the IDE drive solves this problem for writes. Of course with a tranactional system writeback needs to be properly synced. Since IDE doesn't support a fence operation you basically have to turn WB off to guarantee consistancy. For your average workstation or small server user (who doesn't understand all this anyway) its not likely they need that kind of data guarantee, a dropped email isn't the end of the world.
For reads, the outstanding reads don't help much and its more efficient to have the OS do the read ordering because of priority issues. For example a page fault gets moved to the head of the read queue and statified immediatly rather than waiting of existing reads to complete. With the large read caches on the disks grabbing a whole track its not a performace loss to move the head grab a couple sectors and then move the head back to its previous location and read the remaining sectors. The return move will simply be statisfied from the cache.
You seem to have been interested enough to reply. I can't believe a simple post on a message board has got your panties in a wad. Grow up.
"To lead the people, you must walk behind them"
If you use a IDE controller with a I2O processor, this issue is eliminated. Most modern IDE RAID controllers have this feature.
You can get a 15000 rpm drive with scsi, or you can get at most 7200 rpms with IDE.
You now have two identical drives.
Not.
An engineer friend of mine who works at Seagate's Colorado headquarters for IDE drives said that the internal mechanisms for SCSI and IDE at Seagate are different.
The IDE market is FAR more competitive than the SCSI market, so manufacturers have to cut corners wherever they can. Some drive makers put in plastic parts, some put stamped aluminum where steel is usually required, and that is typically why warranties on IDE drives are shorter than their SCSI bretheren.
When you strip the hood off a SCSI drive, you do NOT get an IDE drive with SCSI electronics.
So no. The network is very certainly not the bottleneck.
And on the CPU side, SCSI controllers allow the CPU to sit 95%+ idle while the disks are thrashing their little arses off. I haven't yet seen an IDE interface that good. If you're lucky the CPU will be hit to the tune of 20-50% usage with IDE.
With the 15 devices, who said they were all disks? Oh and the bus runs at 320Mbytes/second.
As I said. SCSI is expensive but designed to scale. IDE is designed to be low cost.
Entirely different purposes, not better/worse.
Government of the people, by corporate executives, for corporate profits.