Slashdot Mirror
Solid State Drives Tested With TRIM Support
Vigile writes "Despite the rising excitement over SSDs, some of it has been tempered by performance degradation issues. The promised land is supposed to be the mighty TRIM command — a way for the OS to indicate to the SSD a range of blocks that are no longer needed because of deleted files. Apparently Windows 7 will implement TRIM of some kind but for now you can use a proprietary TRIM tool on a few select SSDs using Indilinx controllers. A new article at PC Perspective evaluates performance on a pair of Indilinx drives as well as the TRIM utility and its efficacy."
I finally got the opportunity to test out SSDs this year. There may be the odd teething problem to get over, but in my mind there is no market in the future for mechanical drives except maybe as cheap low-speed devices for storing non-critical information... in much the same way as tape drives were used a few years ago.
Which Linux filesystem works best with SSDs? I don't intend to touch Win7.
My rights don't need management.
can someone explain why fragmentation in the mapping between logical blocks and
physical addresses causes performance degradation?
is it an issue with logically sequential reads being spread across multiple pages?
a multi-level lookup to perform the mapping?
?
[perceived] Bottom line SSDs don't work well so, let's just wait until something better comes along.
Also doesn't one of the hardware manufactures (Samsung I think) have a patent on SSD so no one else can make the drives any way. Proprietary == Dead
Why is Windows 7 even in the summary?
People who buy high-end disk drives and care about Windows must be quite a minority. The point of hard disk drives with fast writing performance is for servers.
Well, to be honest, what kind of nerd doesn't like a little trim?
Something as simple as deleting the wrong partition becomes an irreversible operation if you do it using a tool that supports TRIM on TRIM-enabled hardware.
Even if you restore the partition table from a backup, you will likely suffer silent file system corruption, which may even not be apparent until it's too late.
If TRIM support is actually implemented on the device, the device is free to 'lose' data on TRIMmed blocks until they are written at least once.
throw new SuccessException("Sig read successfully");
So does this mean that the a girlfriend of a geek can save her files on it too??
Despite the rising excitement over SSDs, some of it has been tempered by performance degradation issues.
Who cares how they perform. All they have to do is sit there and scare away enemy fleets.
> Actually, magnetic disks have exponentially increased in capacity since the 50s.
> In fact, the rate of increase has been higher than the growth of transistor count.
No, it hasn't!
According to the - slightly stale, from 2005 - article, magnetic storage density has increased by a factor of 50 million in 49 Years, from 2000bit/sq.in. in 1956 to 100Gbit/sq.in. in 2005.
In the same time, transistor count has increased from 1 (single Transistor) in 1956 to a billion Transistors (Gbit Ram) in 2005.
Even if you start in only 1970 with the 1Kbit Dram, you get a millionfold (2^20) increase in just 35 years, or a doubling every 1.75 years, while Disks increased by less than 2^26 in 49 years, doubling only every 1.9 years.
And these days, Hard disks are as dead as a certain parrot!
Unlike in 2005, none of the current must-have gadgets like Iphones, Navigation Systems or high-end Netbooks still sports one, they are increasingly relegated to a role in Grandma's cheap walmart computer holding 10 years worth of crappy photos and videos.
I'd love to get me some trim!
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
I bought a WinTec FileMate Ultra 24G from Tiger Direct that plugs into the ExpressCard Slot. I am now using that as the boot partition with reiserfs (v3), elevator=noop, and mounted noatime. This might not give the very best performance but it is much faster than the stock HD. OpenOffice loads in 2 seconds. I turned down the /sys/block/sdb/queue/read_ahead_kb but I'm not sure where it should be. I put my logs on tmpfs. Some people put the Firefox cache on tmpfs.
The most important point of hard disks are being amazingly multi platform. I didn't like the sound of "Windows driver", "OS support" to perform nicely.
SSD guys really better stick to the standards and never, ever do anything requiring a "driver" on host OS. For example, there are G4 Mac owners who happily upgrades their "old tech" magnetic drives to 500 GB or even 1TB. Who will write driver for them? Apple? SSD vendor? I don't think so.
In fact, HD vendors really better stay away from writing anything except the "smart control" stuff... Or, they better donate to smartctl project and stay away from it too...
Coriolis Systems (who produces iDefrag) jokingly referred to that issue on their blog.
" Ironically even SSDs, where you would expect the uniform access time to render fragmentation a problem of the past, still have various problems caused by exactly the same issue(1)'
of course, they add:
1 For avoidance of doubt, we strongly recommend that you don't try to defragment your SSD-based volumes. The fragmentation issue on SSDs is internal to their implementation, and defragmenting the filesystem would only make matters worse.
In case you spot a good friend who got suggested by Microsoft to defrag their drive (Win7 does it even without asking), you better tell it is not the "magnetic disk fragmentation" issue. It is really different and I heard some real bad stories from people who defragmented (!) their SSD drives.
Am I the only one that read the words "TRIM support" and immediately thought of tight fitting panties?
Will it work on my Commodore?
Program Intellivision!
Apparently Windows 7 will... enough of the promotion of Windows 7 already. I don't object to advertising but I really do hate it when someone tries to pretend that they just want to tell us about one thing, when the real purpose is to tell us about something else.
Where are all the wonderful SSD/HD hybrid drives that were supposed to come out and prevent many of the problems SSDs have?
My dad bought me an SSD for my birthday, and it was one of those models that has horrible studdering issues (which, from what I'm reading, covers most SSDs). That was a lot of money for a drive that let me install drivers about 20 times slower than a hard drive would, and caused my machine to freeze for 20-40 seconds at a time while just surfing the web. That was after disabling all the NTFS caching and optimization nonsense, too.
I remember a long time ago talking to a guy about using a mix of fast and cheap memory in the same computer instead of a huge amount of memory running at one speed. Instead of talking about memory controller costs and OS design, he just told me, "There's no such thing as cheap memory". Well, I seem to see that SLC and MLC have a large difference in cost and performance. Is it at all practical to make a drive with a little SLC and a lot of MLC to aid in performance and wear-leveling issues? SSDs already utilize several megs of cache memory, so is it really that impractical to mix different flash technologies to solve the random-write problems?
I'm not sure I understand the problem. So the SSD thinks it's full but the OS has 'deleted' parts of it so it's really not full, but it hasn't bothered to inform the SSD of this fact, is that the general gist?
OK, so SSD doesn't know that a file is gone and that memory space can be over written.. but then what happens when the OS says, oh, hey, save this new file here please?
How does the SSD know at that point what part of itself it can write the new file to? I can come up with:
1. The OS tells it. OK, so why can't the OS just tell it when it deletes the file and be done with it?
2. It guesses.. somehow that doesn't seem likely?
I have no idea why a TRIM command, which I picture as say, emptying your recycle bin, would necessitate having to move a whole bunch of data around freeing up larger blocks, which I picture as a defrag.
The SSD knows where the files are, or it couldn't access them, it knows where the free space is, or it couldn't utilize it.. so how come it can't simply free up the space on a TRIM instead of moving stuff around?
Is it because each memory space needs to be, hmm, load balanced in a way to avoid early failure? In that case, why not throw more than 1 algorithm at the writes initially, like go up, down, left right, circle, pyramid, etc. (I'm picturing the memory space as a massive field of squares).
What I don't get, is that we have wear leveling in controlers and they also have to translate virtual blocks to physical blocks.
We have file systems that are meant for hard disks, so why don't we have file systems for SSDs? You could just buy any old stupid SSD and get decent performance out of it.
+1 Camel Toe
if by "proven rock-solid" you mean horrid fidelity and media degradation rates, i'd say you are correct about tapes.
[citation needed]
Tapes probably have a better unrecoverable error rate that drives and don't have bits flip randomly while data is at rest like hard drives are found to do. See the talk entitled "No Terabyte Left Behind" given by Andrew Hume at LISA '07 (Wed. 4-6pm):
http://www.usenix.org/events/lisa07/tech/
hey're wasting a LOT of time just retrieving data.
High speed drives can go from shelf to data in a maximum of 60 seconds.
you should really consider a SAN NAS or similar.
Tapes have the highest measure of density when it comes to TB/kW and TB/sq. of data centre floor space. LTO-4 is up to 800 GB native in a tiny little package that takes no power, and LTO-5 is currently in draft and will be 1.6 TB native (add compression for fun). With LTO-4 (and other tape standards as well) and above there's also a standardized way to encrypt the data (AES-256), so if it goes offsite you don't have to worry about data loss.
Tape may not be for everyone, but there are certain things for which there is no replacement for. CERN is using tape to archive the 15 PB/year of data that's going to be generated by LHC: do you want to know how much power to would take to have 15 PB on SAN / NAS? Then take that power and multiply it by 2 or 3 to running the cooling.
It's still a bit early to say SSD has better longevity than a traditional HDD, in theory, yes I agree, SDD probably outlasts HDDs, and this will be even more true as they improve, but I'm not about to jump on the SDD boat 100% just yet.
Better safe than sorry, we've had HDDs for decades and they're fairly acceptable albeit not perfect.
Nerds I'd Love to F#!K
Is there any way to get usage statistics out of the SSD itself? Specifically, I want to query the drive for the number of erase cycles each physical block of flash has been subjected to, verify failures, etc. This should be quite different from the logical block usage at the file-system level, so I don't expect standard HDD tools to give an accurate picture of the drive's health.
Chelloveck
I give up on debugging. From now on, SIGSEGV is a feature.
Thank you. That is a better explanation of the exact problem than I gave, as it explains better why it is so difficult / takes so much time to find a region suitable for erase/overwrite (and why TRIM helps the drive so much).
There's no place I could be, since I've found Serenity...