Slashdot Log In
USB Flash Drive Life Varies Up To 10 Times
Posted by
timothy
on Thu Jun 12, 2008 11:23 PM
from the that's-worse-then-the-spread-for-people dept.
from the that's-worse-then-the-spread-for-people dept.
Lucas123 writes "Differences in the type of memory and I/O controllers used in USB drives can make one device perform two or three times faster and last 10 times longer than another, even if both sport the USB 2.0 logo, according to a Computerworld story. While a slow USB drive may be fine for moving a few dozen megabytes of files around, when you get into larger data transfers, that's when bandwidth contrictions become noticeable. In 2009, controller manufacturers are expected to begin shipping drives with dual- and even four-channel controllers, which will increase speeds even for slower drives."
Related Stories
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
FS (Score:5, Insightful)
Re:FS (Score:4, Informative)
Parent
Re:FS (Score:5, Funny)
Parent
Re:FS (Score:4, Informative)
Mind you it's still free in the sense that you don't pay for it. I'm just annoyed by people using "free software" as a synonym of the business model they favour and expect everyone to know what they mean. Microsoft could claim according to the dictionary that exFAT is free and they'd be right. The FSF doesn't own the word and can't define it. But the exFAT specification is not published (the Sun version of Open Systems) and even if it were the standard would most likely not be an open one in the sense that you don't need a license to implement it (the PC industry criteria for an Open System). Maybe it will be of course, I haven't heard a statement from Microsoft on exFAT openness and licensing.
Parent
Problem with english language, not FSF (Score:4, Interesting)
Latin languages doesn't have this problem, and there's no "free vs. free" ambiguity. Thus nobody speaking those languages has the impression that FSF is playing with words.
Besides, exFAT still costs money to the user, as Vista SP1's license explicitly requires that the user has bought a valid license for Vista, which almost never costs zero, except if the user got it through some channels as MSDNAA.
Parent
Re:FS (Score:5, Informative)
Parent
Re:FS (Score:5, Informative)
Parent
Re: (Score:3)
Re:FS (Score:4, Informative)
I mean USB hard drives. You may not have heard of them, but they exist. I use one.
You evidently did not RTFA. The first sentence is:
The summary does not use the word flash, at all.
I was not saying "RTFA", I was saying: 1) Your assumption is correct
2) The summary should have made it clear so you did not need to make an assumption.
I assume the idiot who modded my comment flamebait also misread it the same way you did. Did you bother reading my comment and the parent properly before replying?
Parent
MARKETING! (Score:5, Insightful)
I want to know SOMETHING like 133x is defined for CompactFlash to give a basic idea of the speed of the device. I'm willing to accept some fudging around but not prepared to find out my new 32-gig flash-drive is 10 times SLOWER than my old 2-gig one. How has this situation persisted this long in a performance-obsessed technical field?
Re: (Score:3, Insightful)
USB thumb drives are like floppy disks, I think no one really cares that much about reliability since most transfers are done via network or CD/DVD/etc, or external storage (portable hard drive).
Re: (Score:3, Informative)
Like Corsair, OCZ or Patriot sticks. If you do your research on the net first then you'll be ok. It's when you walk into a store and they have a selection af candy coloured novelty thumb drives, that's when you're going to get shafted.
Personally I like the Patriot Xporter XT, I use it as a main disk on my NSLU2 debian box. It's not quite as quick as a normal HDD, but it's not bad. Corsair's voyager range are the defacto standard o
Re:MARKETING! (Score:5, Informative)
I tried to use one as the boot drive in my Eee PC and it was glacial. There also seemed to be some kind of pathological interaction between the MCL Voyager GT and Linux's CFQ IO scheduler - when performing a lot of writes the machine would lock solid for several seconds at a time, it looked like reads were being squeezed out. I never did boil it down to a clean test case though. Switching to the deadline scheduler improved matters substantially. While investigating that I realised Linux doesn't have an optimal scheduler for flash drives, they're all built around reducing and consolidating head seeks. no-op (which as the name suggests is just a FIFO with no real scheduling at all) is the fastest scheduler for USB flash, but you get no fair scheduling at all - you have to wait for that 500MB write to finish before your 100-byte read gets its turn. At least some of no-op's better performance is down to it not being anticipatory - it doesn't wait a few milliseconds after an IO to see if the process that requested the previous read/write requests another near by. That's just a waste of time with flash which doesn't have a physical head to seek.
There's a fair bit of tuning you can do at runtime with Linux's IO schedulers, read the docs in
If you want fast, look at the old, 8GB SLC Voyager GTs. 30MB/sec read, 25MB/sec headline figures don't sound that much better, but in the real world they can be 3x faster at writes than the newer MLC models thanks to overwhelmingly better random write performance.
Parent
Holy crap! (Score:4, Funny)
Dual Channel is already available (Score:5, Interesting)
Then I copied a 700 MB file onto it from a local hard drive in gnome, which reported initially that it was transferring at 20+ MB/s, but that dropped steadily until it levelled off around 6.1 MB/s.
Far from scientific, yes, but I wonder a)why the inconsistencies, and b)how these results compare with other products.
db
Re:Dual Channel is already available (Score:5, Insightful)
That's one theory. There may be other reasons.
In terms of how these results compare to other products, I think they are "pretty good". Not the best, but significantly better than average.
Parent
Re: (Score:3, Insightful)
Re:Dual Channel is already available (Score:5, Insightful)
Parent
Proposed rule for Slashdot submissions: (Score:5, Insightful)
Print version here (Score:3, Informative)
Flimsy construction (Score:5, Insightful)
My next drive was a Patriot 2 GB flash drive. It lasted maybe 50 insertions before the usb connector "pushed in" and became so loose that it could no longer be inserted properly into a USB port. I ended up snapping the outer housing off and now it's just a little tiny PCB with chips on it and a USB connector at the end. Works fine but I wouldn't take it anywhere remotely hostile. I keep it next to my computer.
So what is the point of this long story? That flash drives tend to have really cheap construction (in my experience) that doesn't hold up to much use, let alone much abuse. In the case of the Patriot I'm not surprised because it was a really cheap unit. But the Iomega was not.
I don't doubt that th expensive ruggedized flash drives can take much, much more abuse. But they represent like 1% of the market. Most drives are these really flimsily constructed things that fall apart when you look at them the wrong way.
Re:Flimsy construction (Score:5, Funny)
Parent
Re:Flimsy construction (Score:5, Interesting)
This does remind me of what a geek I am though. I think I must be the only person in the world who often daydreams (when I have the time and inclination to daydream) that I've gone back in time and taken some piece of modern computer equipment to shock and amaze people from the early days of computing.
For example, I'll daydream that I've taken my laptop (which is now a few years old and not impressive to anybody, but with 768 MB of RAM, a 40 GB disk, and 1.4 Ghz Pentium M, would have blown the socks off of a computer enthusiast from, say, 1969) back in time and am showing it off to a group of scientists like at say that famous "mother of all demos" where the mouse and graphical interface were first demoed.
Can you imagine showing up and being like, hey check this out. That 64 KB PDP-11 that you have running your demo is cool and all. But let me show you my computer, which has 768 MILLION bytes of RAM! And a 1400x1050 32 bit color flat panel LCD display! With built-in keyboard!
It's not that I would lord it over anyone. But it's fun (for me) to daydream about the conversations you'd have with someone from 1969, explaining to them the advances of modern technology and how they are used in our world.
Anyway, your comment reminded me of that, because although a 2 GB flash drive today is totally ho-hum, if you could sneak one of those back in time to the 1970's, you'd have something that governments would probably go to war over
Like I said, I am a total geek
Parent
Re:Flimsy construction (Score:5, Interesting)
And they would say
You need how many megs to say "Hello World"?
...and I look at the ledger system for the bank where I'm working which uses 3270 sessions. Yes, even though the documentation refers to this as a GUI, hello 1972, your character mode VDU [wikipedia.org] is still haunting us.
Parent
My version doesn't involve scientists, but kids... (Score:5, Interesting)
(And yes, I too have wandered down that sunny Day Dream avenue from time to time.)
You have to be about seven or eight years old, about six months after Star Wars came out and the best computer game available is "Pong" which takes up an entire counsel unit and is still pretty cool. Anyway, you have to show up one day at your friend's house or at school; some place where there's a handful of kids but no adults, and you have to have a modern laptop with you. You explain that it's a new Star Wars toy that they're testing but which hasn't "come out" yet, and that your Dad managed to bring one of the test units home from "the office". Nobody's sure if he really let you have it or if you're going to get in trouble when he finds out. Either way, that's incidental, because everybody is jostling to see what the heck it is you have on the table. It looks like it might have actually been IN a Star Wars film, that's for sure.
Then you crack open the lid and power it up, and muck around with the interface for a while. This should be sufficient to blow your audience away since things like GUI's and mouse pointers haven't been invented yet. --Flat screens which have better color and resolution than any TV set around are also new; just the sort of thing you'd expect a really expensive Star Wars toy to have. You might also want to pull the luminescent CD out of the Star Wars game package and put it in the extending CD tray. --Because Walkman-size consumer electronics have also not been invented, so just the size of the mechanics should also blow your friends away. Not to mention the Buck Rogers CD, (I still think the CD is a dead giveaway that we're all actually living in a low budget sci-fi movie of the week, but anyway. .
Then you start playing the Star Wars game. Music, sound effects, interface, it all looks better even than the best coin-op video game at the mall. A LOT better. You play this for about five or ten minutes, letting your friends have short tries before you suddenly have to go home because your Dad called and you need to bring the game back. And then it is never seen again. Until thirty years later, that is. --The stories which will circulate will not be taken seriously by parents, and yet a handful of kids will be jazzed beyond belief and will be scoping out Department Store Catalogs for the rest of their natural childhoods.
I had a friend who came back from Japan once with a fold-up robot toy which was lightyears ahead of anything our Western toy makers had ever produced. It was one of the coolest days of my entire life. I just picture that day times ten.
-FL
Parent
1980s (Score:3, Interesting)
The mainframe I first worked on did NOT have a card reader, which was odd at the time for the series (NCR Criterion V-series). It had 1MB of memory per CPU, and the two hosts shared a bank of 6 disks, each 500MB (total of 3 GB).
Now, my laptop has a CPU with dual cores, 1GB of memory, and 160GB of disk. Oh, and my cell phone has a microSD card with 2GB on it that is less than my thumbnail size.
Apache with PHP and few PHP script
Re: (Score:3, Funny)
Re: (Score:3, Insightful)
Now I use a 2gb Kingston DataTraveler. It's very solidly built. I use it all the time and carry it with me always. I think it was about $20 last year. They're probably giving them away with a Happy Meal these days. I'm just waiting for the day I lose the little cap.
You can certainly buy very well-built models designed fo
Re:Flimsy construction (Score:5, Funny)
Parent
In my experience(anecdotal of course) physical (Score:4, Insightful)
Linux runs pretty well on my cheap flash drive (Score:5, Insightful)
One annoying thing I have noticed is that programs will periodically completely freeze up and I'll look over and notice that the activity light on the drive is flashing. A common experience is that Firefox will be completely unresponsive, not even redrawing itself when a window that was obscuring it is moved, until the drive stops flashing, and then Firefox will instantly come back to life.
My theory is that the kernel is caching writes to the drive and then at some random point decides it's time to flush the write cache to disk. I think that any program that tries to write any files while the kernel is flushing the cache gets put into a wait state by the kernel until the cache flush is complete and then whatever write the program was attempting, gets written into the cache, ready to be flushed again on the next cache flush.
Furthermore, I theorize that for normal hard disk drives, the write speed is sufficient to keep "ahead" of cache flushes so that the cache never really "fills up" and no programs ever get waited in this way.
But that for slowish flash drives like mine, the kernel doesn't compensate for the slow write speed of the flash (because the kernel doesn't even realize that it's writing to flash?) and so it lets enough data buffer up that it has to frantically try to flush it all when the cache has filled up. Or perhaps, that the kernel just tries to flush too much at once, not realizing how slow the flush is going to be due to the underlying speed of the device.
I also theorize that this problem could be solved by having the kernel flush the cache more aggressively, and in smaller increments. If the flash drive were kept continually busy flushing small chunks of write cache, then a) the write cache would not be as likely to fill up, and b) no individual write would monopolize the device for such a long period of time becase the writes are all smaller.
Writing all of this makes me realize that the root cause may be that programs are trying to *read* from the device while a write cache flush is happening, and since the device can only do one operation (read or write) at a time, the long duration of the cache flush operation is blocking a program from reading the drive. Furthermore, if what the program is trying to read is a demand-paged part of its text segment, then it makes perfect sense that the whole process would be blocked by the kernel while the text segment piece waits to be loaded.
Am I even close to the mark on this one?
If so, I am sure there are Linux kernel experts who can tell me what values to write into what
Pauses when running from flash drive (Score:4, Informative)
db
Parent
Re: (Score:3, Interesting)
db
Re: (Score:3, Insightful)
Re: (Score:3, Informative)
Re:Linux runs pretty well on my cheap flash drive (Score:4, Informative)
I touched on this in an earlier post, but I experienced the same thing and tracked it down to the CFQ IO scheduler. I never got round to making a clean test case so never submitted the bug to the kernel. My wild-ass guess is that it's assuming reads and writes take equal time when they don't with flash, which confuses its notion of "fairness". It's a pity CFQ doesn't seem to work well with flash, as it has some tasty features like being able to ionice your backup process so foreground tasks get priority. It's not just me who things CFQ aint great for flash - the default kernel on an Eee PC doesn't even have it compiled in, despite CFQ being the default scheduler for the kernel version they use (it's still the default now, you're almost certainly using it).
The IO scheduler operates at the block device level, below the page cache daemon, so in theory even when dumping cache it shouldn't starve your reads and writes out. That's pretty much the point of having a clever IO scheduler - not having to wait for a 5 GB IO to finish before you get to read the 100 byte file you're blocking on. Mostly that's what Linux's schedulers do. But CFQ with a flash drive? Not so much.
You can change the IO scheduler and tune it at runtime, which is very handy indeed. This shouldn't cause data loss, and I've not had any problems.
I'll assume your flash drive is /dev/sda, change sda as appropriate.
To see which IO scheduler you're currently using and those available in your kernel:
That shows I have all four schedulers available and CFQ is currently in use. To change it, just echo the name of the new one to that file:
I found the deadline scheduler had much better interactive performance that CFQ when booting from a flash drive.
Changing the scheduler as detailed above needs to be done after every boot and for every device. If you want to use a particular scheduler as the kernel default for all devices, either choose it at kernel compile time or pass the "elevator" parameter to your kernel in your bootloader config. For example, "elevator=deadline" makes the deadline scheduler the default. If you use grub, tag that on the end of the kernel line in /boot/grub/menu.lst.
There are tuning "knobs" for each scheduler. Read the docs in /linux/source/Documentation/block for the gory details. I tuned my system to use the deadline scheduler, group IOs less (no head seek penalty) and prioritise reads over writes (things regularly block on reads but less so on writes, which can be cached anyway - remember we're working below the level of the cache). Read the docs to understand the gory details - note that these tunables are for the deadline scheduler, they're different for different schedulers. They're not very scientifically selected and not exhaustively benchmarked so you may be able to do better. The numbers for expire are in milliseconds. This lot will dispatch IOs individually (instead of in groups of 16), do 10 reads for every write, prioritise a read a
Parent
Thumb Drive with flexible neck (Score:5, Interesting)
Re:Thumb Drive with flexible neck (Score:4, Insightful)
Parent
Brand name is no guarantee (Score:5, Interesting)
Re: (Score:3, Informative)
a) Sony Tiny 2GB: 6.2 MB/sec
b) SanDisk Cruzer 2GB: 7.1 MB/sec
c) Patriot Xporter 2G: 14.2 MB/sec
d) Patriot Xporter 8G: 11.7 MB/sec
e) PQI 4GB: 1.5 MB/sec
a & b seem like "typical" drives. c was supposed to be fast, and it is! I liked it so much I bought d. Drive e was a big mistake - impulse buy without knowing the specs. It is too slow to use (45 minutes to fill it up.)
Alas, the Patriot 2GB just went in under RMA yesterday. It wa
Re: (Score:3, Insightful)
Re:What does speed have to do with life? (Score:4, Informative)
Parent
That's easy (Score:5, Funny)
Parent
Re: (Score:3, Funny)
Re:Flash MP3 player (Score:5, Interesting)
Parent
Re: (Score:3, Insightful)
Do you have objective numbers to back up that claim, like perhaps some RMAA tests? Or are you acting no better than the subjectivist audiophools who spend thousands on power cables?
Re: (Score:3, Insightful)
Re:I thought everyone knew this (Score:5, Insightful)
Sure, you get what you pay for. But the problem here is that these drives don't indicate on the packaging whether they use SLC or MLC memory, or whether they offer more than one channel.
So let's say the crappy variety of 4 GB USB drives currently go for $25, and the better, faster variety will never sell at that price. Right now, you have no way of knowing whether that 4 GB drive going for $50 is made with the faster, more durable SLC memory, or whether the drive is simply overpriced.
You can therefore spend $50 for a drive and not get what you paid for. And the only way to safeguard yourself is to waste time researching your drive -- something you shouldn't have to do, since this info ought to be published as part of the drive's specifications.
Parent
Re:I thought everyone knew this (Score:5, Informative)
Logging is the Flash RAM killer.
And Kingston and Sandisk should start putting "SLC" or "MLC" on their products, so we techies know whether they are worth the double price.
Parent
Re:I thought everyone knew this (Score:5, Informative)
I was just discussing this the other day, and my friend found this: http://www.kingston.com/ukroot/flash/flashendurance.asp [kingston.com]
Parent