Should I defrag a SSD (Crucial M4)

[Deleted]

This is definitely wrong.

Defragmenting is done like this:
A) copy the fragmented file to a free place in the disk
B) Move away any other files in that old place until we have enough free space to fit the first file
C) Copy back the first file to its old place. As the other files are gone, it fits into the free space completely, rendering it unfragmented.

As you can see we have a lot of write operations here and these operations wear down a SSD. Remember: A specific cell can only be written abount 10k times. Defragmentation reads and writes files very often over the same place, aging this cell rapidly. If the cell is gone, the whole chip is automatically disabled, rendering the SSD alsmost completely unfunctional.

Thats why we always state: Never defragment a SSD. Its useless AND it kills the SSD.

[Snorkle]

5-6 years unless you are running it in a server or a RAID configuration that is constantly shuffling data around. Longer than the useful lifetime of the computer at any rate.

It used to be that way. However, the type of MLC NANDs being used by the newest generation of SSDs have a shorter lifespan, sometimes of up to 10 times shorter. You should do some research.

[Butler to Baby Sloths]

It used to be that way. However, the type of MLC NANDs being used by the newest generation of SSDs have a shorter lifespan, sometimes of up to 10 times shorter. You should do some research.

They are really selling SSDs with an expected lifetime of only 7 months? Somehow I find that hard to believe, especially as they warrant for 24 months.

[vesseblah]

They are really selling SSDs with an expected lifetime of only 7 months? Somehow I find that hard to believe, especially as they warrant for 24 months.

I tried to educate him two weeks ago on the same subject, but no hope... I recommend not even trying.

[Snorkle]

They are really selling SSDs with an expected lifetime of only 7 months? Somehow I find that hard to believe, especially as they warrant for 24 months.

2 to 3 and a half years if people use them properly (moving away swap, temporary files, turn off superfetch and prefetch, use TRIM, etc etc etc).

You can read several white papers and independent tests as well as manufacturer tests on the issue, and vesseblah, your "education" was completely ignoring several links proving you wrong. Last I heard, education was sharing knowledge, not denying it.

[Mythricia]

2 to 3 and a half years if people use them properly (moving away swap, temporary files, turn off superfetch and prefetch, use TRIM, etc etc etc).

You can read several white papers and independent tests as well as manufacturer tests on the issue, and vesseblah, your "education" was completely ignoring several links proving you wrong. Last I heard, education was sharing knowledge, not denying it.

I've been using an older generation SSD and I'm a pretty much a power user, so it's done a ton of work. And it's still going, performing great, and is older than that. So, really, people saying SSD's only last 2 years or what not is bogus, and people saying it broke before that has either; used it wrongly, intentionally worked it much harder than even a power user would let alone a normal average user, or had a faulty unit from the get go.

In practise most SSD's will far outlive their usefulness.

HDDs break too, used intensly enough. SSD's are no different.

[Deleted]

You should do some research.

Current flash cells can handle somewhere in the region of 3500 cell writes before failure. Thats approximately 420TB of data written on a 120gb drive before failure. Theres write amplification after that however, lets assume 2x (a very high number). Thats 210TB of realistic use data.

My own usage puts approximately 2TB of data written to the SSD per year. Others could easily have lower write rates than that. The SSD thats used primarely is currently nearing 3 years of age, and possesses only 4 reallocated sectors. (3 of which showed up in the first 12 months)

SSD retention (how long any specific data sector can hold before it resets) can be anywhere around 1-4 years, but a reset cell is still usable. (This, along with capacity, is why SSDs make bad storage/archiving drives)

[Snorkle]

I'm not gonna argue, I've presented SEVEN different links with info, 2 university studies, 2 manufacturer papers and 3 independent tests. If that's not enough, I'll leave people to their ignorance, but I do have one very important question:

Why do people think "failure" means "every single NAND is broken" ?

Wait I have another one:

Why do people think "the new MLC NANDs used in the newer generation of sandforce controlled SSDs, et al" means "every single SSD in existence"?

If I had the answer to both those questions, maybe all of this made more sense to me, but alas, I do not, and as such, this makes NO SENSE.

[Butler to Baby Sloths]

In the post I replied to you neglected to mention that you were only talking about sandforce SSDs. It read like you were talking about every new SSD.

[Deleted]

Why do people think "failure" means "every single NAND is broken" ?

Generally a 'drive' failure means the drive is completely inoperable.

If you're talking about controller failures, then thats not an SSD specific issue. HDDs have plenty of controller deaths as well, as you should already be aware of. (They're just in context more difficult for the average user to tell apart from a mechanical failure)
Sandforce is notorious for dying or having issues, but this has been the case with even the first gen controllers (which are not present in 'current' SSDs).

Why do people think "the new MLC NANDs used in the newer generation of sandforce controlled SSDs, et al" means "every single SSD in existence"?

You did not once mention Sandforce or only that its specific models of current SSDs, many SSDs are not Sandforce based, and many Sandforce SSDs vary which flash storage they use. The M4 uses the Marvell 88SS9174-BLD2 controller, so again sandforce is not relevant to the topic drive.

I've presented SEVEN different links with info, 2 university studies, 2 manufacturer papers and 3 independent tests

I've seen multiple papers bringing up the issue of data retention - and the reason I brought that up is because you've not presented your links. Which means I cannot be certain whether or not you're confusing data retention time with drive lifetime/durability. Several users have already done so in the past.

[Deleted]

Since there seems to be some controversy about this I thought it would be helpful to give a detailed explanation:
You should never defrag an SSD. Don't even think about it. The reason is that physical data placement on an SSD is handled solely by the SSD's firmware, and what it reports to Windows is NOT how the data is actually stored on the SSD.
This means that the physical data placement a defragger shows in it's fancy sector chart has nothing to do with reality. The data is NOT where Windows thinks it is, and Windows has no control over where the data is actually placed.
To even out usage on its internal memory chips SSD firmware intentionally splits data up across all of the SSD's memory chips, and it also moves data around on these chips when it isn't busy reading or writing (in an attempt to even out chip usage.)
Windows never sees any of this, so if you do a defrag Windows will simply cause a whole bunch of needless I/O to the SSD and this will do nothing except decrease the useful life of the SSD.

[lockedout]

No more damage than making a step actuator fly around a platter drive for two hours moving shit around. -_-

Defragging won't cause damage. It just won't do anything meaningful.

When you defrag a hard drive, especially a SSD, it will end up causing tons of small write accesses, which will reduce the lifespan of the hard drive. And since data is not being read sequentially, it doesn’t matter if the file is stored in a hundred different places, the performance will remain the same.

So, no, you should not defrag an SSD. And performing one will actually reduce the life of your drive. All of the SSD manufacturer’s know of this problem and they have come up with an optimization technique with the use of the TRIM command.

[alexklein]

While traditional defrags should never be performed on an SSD, these drives, while they offer a faster and better experience compared to standard HDDs, will over time experience a very noticeable and dramatic deterioration in performance. The company I work for, < snipped >, makes software called < snipped > that optimizes SSDs with technology called < snipped >. Have any of you tried < snipped > before?

-Alex

[Deleted]

Lol SSD's have such a fast access time I highly doubt defrag even matters.

Offtopic, why is everybody whining about the fact that SSD's have a "short" lifespan and they die after a couple of years. HDD's die aswell, they also have limited writes.

[TheSerialSniper]

SSD's last much longer than HDD (no moving parts that wear and tear) and even when they do go up it is MUCH easier to retrieve data off of them than HDD's

[Milkshake86]

Defrag is for mechanical drives, what it does is basically sorting the data on the disc in order of how much it is used so that data you use often is closer to the middle where a mechanical drive is at its fastest.
In a SSD there is no slow or faster parts so it would do nothing for you except using a few of the limit writes a SSD can handle.

This aswell as an SSD stores things in order as is, it doesnt drop stuff around.