SSD wear meme

>>57728531
>being capable of even doing this because your SSDs aren't in a RAID

>>57728563
>using more than one SSD

literally retarded

File: ssd.png (83KB, 688x960px) Image search: [Google]

83KB, 688x960px

is mine too old?

File: Boot Disk.png (40KB, 866x746px) Image search: [Google]

40KB, 866x746px

>>57728660
lol ok

Is there a tool for this on Linux?

>>57728798
>i run linux
>but am too retarded to figure out how to query SMART attributes
why do you fags even run linux if you cant google something simple or use the cli

File: 2016-11-28_17-02-18-snip.png (1KB, 297x50px) Image search: [Google]

1KB, 297x50px

>>57728798
Here borrow mine, no warranty: http://ix.io/1IAn

Samsung 850 Pro 256GB - 3,864 hours on.

>>57728949
>why isn't linux more popular?
>fucking casual user, why do you even use linux

>>57729178
>why isn't linux more popular?
I never asked this. The only reason linux is popular is the lack of licensing fees. And the only reason it is popular on consumer devices is when it has been dumbed down so hard that you have to void your hardwares warranty to access the CLI like on android phones.

>>57729178

That makes the first statement akin to "Why aren't there more people in the world who aren't incompetent?" which is totally valid.

File: clip+(2016-11-28+at+01.36.50).png (62KB, 674x532px) Image search: [Google]

62KB, 674x532px

>>57728531

File: a.jpg (55KB, 668x207px) Image search: [Google]

55KB, 668x207px

>tfw my msata ssd is 23 degrees celsius above the fucking CPU
>it's not even running at sata3 speeds because my motherboard a shit
>tfw it's going to explode because samshit
Why the fuck did i listen to the Samshills

>>57728798
gsmartcontrol

File: Capture.png (60KB, 674x532px) Image search: [Google]

60KB, 674x532px

How to total writes?

Or is it because the controller on my ssd

File: 1476222329507.png (560KB, 1002x692px) Image search: [Google]

560KB, 1002x692px

I'm fine, right?

Bought it December 22nd last year

File: lel.png (190KB, 1555x839px) Image search: [Google]

190KB, 1555x839px

>>57729708
in the same boat, anon
I audibly kek'd

File: 1468413871839.png (26KB, 671x201px) Image search: [Google]

26KB, 671x201px

>>57728642
>>57729248
>>57729294
>>57729365
>>57729717
Are you people even using the thing?

>>57729793
We treat them like our first-born. So of course our SSDs are coddled.

>download this
>my seagate external 2 TB has been on for 3 years of use
>still good health

Damn.

>>57729793
Only the OS and Overwatch is installed.
Should I bother putting other stuff in it? Cubase, VSTs and other games?

>>57729952
I have all of my programs and do all my development on that ssd.

File: Screenshot_2016-11-28_02-10-52.png (115KB, 954x742px) Image search: [Google]

115KB, 954x742px

>tfw Linux doesn't have a Shizuku-chan

File: ssd.png (73KB, 674x532px) Image search: [Google]

73KB, 674x532px

>>57728531
got it april 2015

File: mm.jpg (165KB, 1000x690px) Image search: [Google]

165KB, 1000x690px

Memetez 3 still going strong.

File: Capture.png (369KB, 1252x859px) Image search: [Google]

369KB, 1252x859px

>>57729995
Why even live?

>>57729158

>44 years remaining

yeah right fag. you yourself won't even live that long.

File: 123.png (70KB, 680x761px) Image search: [Google]

70KB, 680x761px

>>57728531
desuu~

>>57730179

It's there to serve as a reminder that I should stop being so conservative with SSD writes.

You're right tho, it's closer to 200 years, but the lifetime % smart value only reads 99% instead of a more accurate ~99.8%, assuming a very conservative estimate of 2000 total rewrites / 512TB written (same type of SSD withstood 15x that in a stress test).

Model : SanDisk SDSSDHII240G
Disk Size : 240.0 GB (8.4/137.4/240.0/240.0)
Power On Hours : 10067 hours
Power On Count : 775 count
Host Reads : 20030 GB
Host Writes : 14244 GB
NAND Writes : 8349 GB
Health Status : Good (100 %)

>>57729793
You have more hours clocked than them.

File: DiskInfo64_2016-11-28_12-37-40.png (82KB, 674x652px) Image search: [Google]

82KB, 674x652px

SSD wear is a meme. Been fucking my SSD in the ass forever and it still runs without any flaws.

>>57731770
wew lad

File: SSD.png (539KB, 1036x713px) Image search: [Google]

539KB, 1036x713px

Who's comfy here in winter. SSD below 30c

>>57731770
What did you do to get 134TB write?

Has anyone here actually had an ssd fail yet? Like, even when they fail, they just go read-only so who gives a shit?

>>57732818
In those controlled write tests, they usually brick after they reach their limit. But only after obvious tell-tale signs show up (many rellocated sectors).

File: ssd.jpg (66KB, 401x363px) Image search: [Google]

66KB, 401x363px

>>57728531
I fell for the SSD meme, I'm going to stick with mechanical hard drives from now on.

File: file.png (98KB, 674x656px) Image search: [Google]

98KB, 674x656px

>>57728531
3 years running.
I also have a retired 60gb SSD from the time they first really hit the market, still runs fine with total hours and writes higher than this one.

File: Screenshot from 2016-11-28 15-11-24.png (39KB, 553x301px) Image search: [Google]

39KB, 553x301px

Come at me.

>>57728531
Not at home rn, but:
250GB Capacity (Samsung 840)
A little under 24,000h power-on time
13TB of writes.

Health reported as "Good" in Samsung Magician.

File: Capture5454.png (81KB, 670x249px) Image search: [Google]

81KB, 670x249px

Should I start panicking?

>>57731770
Plextor master race

>tfw can't see my ssd writes/reads

>>57730729

>it's closer to 200 years
>ahahahahahahahahaha

You SSD cucks are in full denial. The speed decreases 20% in just the first 6 months, SSDs are throwaway storage units.

File: ssd.jpg (300KB, 1002x814px) Image search: [Google]

300KB, 1002x814px

>>57728531

>>57734837
>The speed decreases 20% in just the first 6 months
>what is TRIM
how was your time machine trip? which year are you from?

>>57732718
Most of it is from the many database conversions I deal with.

File: 1455122603501.jpg (32KB, 375x344px) Image search: [Google]

32KB, 375x344px

>>57734890
Heck, there's SSDs with automatic garbage collection so the user doesn't have to fuck around with it anymore.

I have a Crucial M4 from 2012 that I put in my sister's macbook which she's used everyday. I can only imagine what it would say in crystaldisk. I'll have to check it out sometime.

File: FB_IMG_1460967523083.jpg (14KB, 378x293px) Image search: [Google]

14KB, 378x293px

SSD related question:
If I allocate ~90% of my SSD (let's say 100 of 120-128 GB), will the remaining (unallocated) part be used for TRIM and other optimisation n' stuff?
I have no need for the extra space, and I like to keep my partition sizes round.

>>57734890

>what is TRIM

Something that when used as a non-queued command takes a lot of time.
Something that when used as a queued command causes severe data corruption.
Something that you can use to tell where is data and where isn't in an encrypted volume.
Something that doesn't work with Linux.
Something that still doesn't prevent the 20% speed reduction in the first 6 months.

which year are you from?

Obviously older than you.

SMART WRITE LOG does not return COUNT and LBA_LOW register
SCT (Get) Error Recovery Control command failed

Device Statistics (GP/SMART Log 0x04) not supported

ATA_READ_LOG_EXT (addr=0x11:0x00, page=0, n=1) failed: 48-bit ATA commands not implemented
Read SATA Phy Event Counters failed

rip OSX, it only shows my 12351 hours.

>>57735081
You don't need to leave free space anymore. Pratically all SSDs nowadays have some hidden space for that.
>>57735206
You're so full of bullshit. First of all, you don't even need TRIM anymore because nowadays most non-crappy SSDs have controllers with proper Garbage Collection algorithms.

>Something that when used as a non-queued command takes a lot of time.
It takes a few seconds on SSDs with GC and max. a couple of minutes on bad SSDs, what the fuck are you talking about?
>Something that when used as a queued command causes severe data corruption.
Yea, maybe if you have a crappy Samsung SSD because they can't into Firmware.
>Something that you can use to tell where is data and where isn't in an encrypted volume.
https://vxlabs.com/2015/02/11/use-the-hardware-based-full-disk-encryption-your-tcg-opal-ssd-with-msed/
>Something that doesn't work with Linux.
fstrim
>Something that still doesn't prevent the 20% speed reduction in the first 6 months.
it does, you are stupid. is that what you always do even though you have no clue about this subject? throwing technical terms at people ?

>>57735206
>>57735337
I am >>57731770 and my SSD runs just as fast as when it was new.

>>57735337
I know I don't need to do it, I'm just asking if that's 100% pointless if I will do it anyway.
What's the point of having "111.47 GB" while I can make an ideal 100GB one (and still use maybe half or less)

>>57735378
>I'm just asking if that's 100% pointless if I will do it anyway.
What kind of SSD do you have?

>>57735406
several:
-120 gb for os
-256 for data
-512 for games
same as above but w/o 512 in laptop
a few other lying around and serving me as portable drives

>>57735406
>>57735518
I'm dualbooting btw, so the os ones are split in two (50+50)

File: SSDLife.png (53KB, 552x767px) Image search: [Google]

53KB, 552x767px

It simbly twerks

>>57735518
>>57735530
Sorry, wasn't I precise. I was asking for the brands/models.

My XLR8 was good for about 2 years then the write speeds on it plummeted significantly. One day it was fast as hell then the next it was slower than my regular hard drive. I did a secure erase on the SSD with parted magic and it didn't restore any speed to it.

File: 1314340141299.png (87KB, 400x400px) Image search: [Google]

87KB, 400x400px

>>57735535
>OCZ

>>57735550
Oh, sure.
samsung 850 256 gb (pro in desktop, evo in laptop)
kingston hyperx3k 120 - boot drive in desktop
corsair hs 120 (not sure ?) - boot drive in laptop
crucial cx200 500 gb - "gayming drive" in dekstop
one more hyperx3k, one more crucial (not cx200 tho, something older I believe) as spare drives
That's it, I believe.

I'm mostly working with data containers located on my "data" (256G) drives, boot drives are just there to keep OS alive

>>57735659
ikr? Still just werks.

>>57735742
All of your drives have proper controllers, there's no need to worry about the 100% thing or actually anything.

>>57735947
I correct myself since you asked for it: Not only don't you have to worry, it's actually pointless to leave free space.

File: Capturar.png (794KB, 998x636px) Image search: [Google]

794KB, 998x636px

>no money to buy a new one
>external is full
How fucked am I?

>>57735947
>>57735963
Good to know, thanks.
Anyways, I'll probably still keep making the largest round-size partition possible and leave the rest unallocated.
fashion > stats

>>57736031
Is it a backup or the only set of data you have?
I have a Seagate Barracuda 7200.11 1.5tb which is notorious for failing. Out of the 4 I started with one remains and it's going stronk even with 24 reallocated sectors. I have backups though so I'm not worried.

If this is the only copy of your data then find money for backups or be prepared to lose your data either in the next few hours or months. It's a gamble.

>>57736031
get used one for a literally few bucks and backup all your data
2 ticking bombs have rather small chance to explode at the same time

>>57732899
Nice chinkshit garbage

>>57732899
>I defragment my SSD

>>57736158
Yeah, the only backup I have is for wallpapers on the cloud and my 3DS NAND. Guess I'll watch some anime in order to get space on the external and shove the important stuff there.

>>57736213
I havent trusted a single hard drives for over a decade now. Doubling everything has saved my ass multiple times. I know it's more expensive but either you can afford to lose your data or it's simply not important enough to warrant buying a second drive for.

>>57728531
NVMe consumer SSDs are a meme.

Can anyone recommend good ways to encrypt a home folder on an SSD? I hear using the standard ecryptfs option is a bad idea for SSD and causes issues with TRIM. Is setting a BIOS password on a Samsung 750 Evo enough to do the business?

File: wdg.png (113KB, 667x689px) Image search: [Google]

113KB, 667x689px

I don't know what any of this means.
It's a WD Green that came in a prebuilt that I've had for seven years now, used daily.
Is this bad? Should I get a new drive?

>>57738951
Your SSD does not have the F1 attribute that keeps track of the total number of bytes written to it.

>>57739043
It's not even solid state.

>>57736270
That's good and all but it has become an obsession to me, i backup everything on 3 different drives,
that means that when i fill 4tb of data i need to buy 3 more 4tb, and believe me i do fill them, and they are all eromanga eroges and jav, goddammit.

>>57739059
It's bad. Throw it out the window before it explodes.

Actually no.

HDDs don't track the number of writes because they have an infinite number of write cycles. CrystalDiskInfo shows you all SMART attributes in hex. All the important attributes are at 0. You're doing fine.

>>57728531
How can I see this information on OS X?

>>57730076
Cannot into /dev/sd{a,b,c}?

>>57732818
>even when they fail, they just go read-only

Except that's wrong, retard.

>>57734791
It must be tough being mentally retarded.

>>57739156
Thank you. I might still pick up a 3TB drive for $70 on sale, just in case.

>>57739100
Like I said either it's important or it isnt. I have about ~22tb raw storage which is just under 11tb usable space. Are you running enterprise rack gear yet?

>>57739100
Don't buy 4tb drives. 3tb Toshibas are way cheaper than anything else.

>>57740041
How do Toshibas rate against WD and Seagate? No one ever mentions them.

>>57740058
I can either get a 4tb 5400rpm wd red meme drive for £130 or a 3tb toshiba 7200rpm for £70. I regret falling for the Red meme and should have instead bought some used 12bay DAS and bought 12x3tb Toshibas.

Most HD manufacturers have had slip ups. The only drives to fail on me so far are Seagate and WD but I've also owned more WDs and Seagate's over the years too.

I say if you don't care about performance too much then find pretty much any drive you want, if you want some performance go for the cheapest 7200rpm. Don't fall for meme labels like NAS or Green or whatever fucking bullshit they try to attach to it to inflate the price.

I don't know what any of this means

Is it good or bad? I've had the drive for around 5 months now

File: Capture.jpg (140KB, 659x819px) Image search: [Google]

140KB, 659x819px

>>57740393
Fuck forgot image

>>57740393
>>57740407
>GOOD
>100%
>wat does it mean guis?

Apply yourself

>>57740513
Kill yourself.

How long does an SSD last assuming daily use? I can afford a relatively big one.
A quick google search says 10+ years under stress tests, but that doesn't sound right.

File: 658153.png (531KB, 1016x696px) Image search: [Google]

531KB, 1016x696px

Had it for two years so far, but I try to only use it as a boot drive.

SSD newb here, I see people shittalking Samsung. Should I get Crucial instead?

File: magician.jpg (161KB, 1024x728px) Image search: [Google]

161KB, 1024x728px

Seems like I write about 1TB a month

400TB endurance is enough man, by the time I am even close this will be long obsolete or replaced

>>57745781
Seconding this. About to pull the trigger on a Samsung, can someone talk me out of it?

File: smartctl.png (26KB, 839x1525px) Image search: [Google]

26KB, 839x1525px

>>57745781
I vote samsung. Don't have any reason to trust crucial

>>57746926
>old age
>pre-fail
What does it mean

>>57745781
>>57746906
If you're getting a SATA SSD, you probably won't be able to tell the difference between the high end SSDs of any reputable brand. Samsung 850s are great, but other brands have drives that are just as good. At this point, the SATA interface is saturated by most drives. That said, there are also "budget" lines like the Samsung 750 or Kingston UV400 that won't saturate SATA, but are still much faster than HDD and are great for a data drive if you don't have massive needs (maybe have NAS) .

>>57747024
“Old_age” type means it's a metric that's designed purely for tracking age, not necessarily indicating any error

For example “number of hours powered on” and “crc error count” are both things expected to go up naturally at a consistent rate as the SSD ages

“Pre-fail” means it's a metric that's designed to track errors indicative of near-term or imminent failure, e.g. “reallocated sector count” - once you start seeing those you definitely want to have a replacement drive on premise

Are Toshiba P300 3TB hhd's okay for everyday use?
I used to be told anything more than 1TB would fail quickly.

>>57747151
This is an SSD thread, noob, take your pleb shit somewhere else.

>>57747133
What Raw_Value do you want next to those metrics, then? Because I'm seeing 0s and a 97 in that image. Does that mean 20k hours is all you'll get out of a Samsung?

>>57747171
s-sorry

>>57747200
You better be.

File: vitals-840pro-life.gif (9KB, 618x326px) Image search: [Google]

9KB, 618x326px

>>57747180
Most of the time, raw_value is the only stat actually worth caring about IMO. The translation from raw value to “percentage” is mostly arbitrary.

In some cases, there's not even a sensible translation you could perform at all (e.g. in the case of Airflow_Temperature), so only the raw value is relevant. That's especially the case for something like Power_On_hours.

>Because I'm seeing 0s and a 97 in that image.
For those percentage conversions, lower = worse. So a new drive would have 100% everywhere in theory. (Don't ask me why ECC_Error_Rate is at 200)

A good example of a metric where the percentage actually makes sense is Wear_Leveling_Count and Used_Rsvd_Blk_Cnt_Tot. Both of those are basically an indication of how much of your SSD's write cycles are still remaining. On my drive, I've used 97 P/E cycles (wear leveling count), and the drive is rated for 6000, so it displays it as 5903 / 6000 = 98.4% rounded down to ‘98’.

Pic related is the lifecycle of an 840 pro put through an endurance test. First all of the wear leveling is exhausted, meaning blocks start becoming unusable, so it has to get spare blocks from its reserved block storage (i.e. Used_Rsvd_Blk_Cnt_Tot would start going up) to compensate. Only once you've exhausted the spare block reserve would you expect an SSD to die from write exhaustion.

In my particular drive you can see that this would not be the case even in 100 years of continued operation at my current usage rate. In summary, the drive looks extremely healthy and I expect 10+ years of service at least out of it, although I'll probably replace it by a faster and larger NVMe/PCI-e SSD by then.

File: mx100.png (235KB, 672x292px) Image search: [Google]

235KB, 672x292px

The health drops so fast when I write to it...

>>57747790
>11% health loss from 70 P/E cycles
holy shit that drive is either ancient or terrible

>>57747803
It's the damn Crucial MX100 /g/ was meming back in the day.

>>57747790
>>57747812
I don't trust the health bar anyway. So I won't worry.

>>57747706
(also note that this graph shows the 840 generation, which was only rated for 3000 P/E cycles. The 6000 P/E cycle figure is for the 850 generation, ergo it would probably last twice as long in practice)

File: Intel.png (53KB, 674x696px) Image search: [Google]

53KB, 674x696px

Bought and installed mine sometime in 2012, used for my OS. Games are installed on the HDD

What does 91 mean for "Media Wearout Indicator?"

Also
>Unsafe Shutdown Count
I didn't know that was recorded
>BE
Shameful