It died because it's an OCZ goddamn Vertex. Actually it's miraculous that it even survived this long. SSDs rely heavily on internal firmware to function correctly day in and day out, and OCZ spent years shipping terribly buggy firmwares that would eventually brick most of the drives. Some sooner, some later, but enough failures to drive a previously respected parts company out of business entirely.

That said, all drives can die at random for any number of reasons. Be prepared.

Always surprised to see how much hardware fails for other people. Do you guys have... power spikes or something? Or do you live in an arctic environment, place the PC on the heater and regularly open a window, or something? Or are your disks mounted in a hummer on a military mission in Iraq? It seems like I never see that kind of thing.

I have OCZ Vertex from around 2012 (possibly 2011?) and OCZ Core from around 2008-09 still plugged into my computer because they work and I'd rather use them for the page file and temp folder than not use them at all. They've survived one and two PC upgrades, respectively. Did that firmware upgrade which allegedly kills the disk too (I only read about it after doing the upgrade). Nothing.

Using a Samsung Evo840 as system drive for several years too (not sure, 2013?), and no problems. I replaced the Vector with the Evo because reviews said it was so much faster (and that's true), not because it made problems or ceased to work alltogether.

Although of course I've had my first harddisk in 1985, and Samsung is the only ever company to produce a harddisk that has failed me (I've had some Quantums and Seagates up and running literally for decades)... so I kind of half-expected that Evo840 to fail me. But no, nothing.

The 5 Seagates in the NAS here are up and running 24/7 for two years too, recording security footage at night and scrubbing ca. 9 hours daily during work time. No fails, no errors, and SMART metrics are all virgin. I wouldn't expect anything different either. If one of them failed me after only two years, I'd send some nasty hate mail to Seagate. No, they're not enterprise-grade disks, simple "ST1000" Barracudas.

If one of them failed me after only two years, I'd send some nasty hate mail to Seagate. No, they're not enterprise-grade disks, simple "ST1000" Barracudas.

Exactly. Platter based hard disks often have a five year warranty now. Do ssds have the same? I expect my disks to run for decades without failure too and generally they do...

The fact that an ssd failure usually takes out the whole drive spectacularly and bricks it, whilst a platter hard disk failure might just ruin a couple of sectors, to me is a huge reliability flaw. At least with a non ssd if you get a failure of a few sectors, smart flags it up and you have some time to go out and buy a replacement rather than being without a pc until you can get a new disk from Amazon or whatever, and what's more you might even be able to salvage the disk image to the new drive saving yourself time, as it can take an age to restore a backup from a cloud backup service...

In a way this reminds me of the very first moving parts hard disks. They were a bit unstable and would crash if you even looked at them funny. You even had to manually put the disk heads in a park position before you moved the disk somewhere physically different or the heads would scrape the disk platter destroying the entire surface. Ouch.

In that case and with ssds I think perhaps users are still early adopters and are taking that risk...

I too never managed to have a total HDD failure, although experiencing CRC error with my Seagate ST2000VX (so called continous write security grade) when there is space less than 50 GB left.

still I'd wish there was a viable solution to back up 2 TB of data.

HDD builds up slowly, starting with bad blocks, and it can be quite some time before a failing disk is actually unusable. SSD tends to die in style; it goes out instantly.

That's actually a very important remark you did.

HDDs have lots of signs they're about to die, because they're mechanical. SMART can for example monitor the power draw. An HDD that is about to fail is probably drawing substantially more power (or with intermittent spikes), having a higher than normal spin up or spin down time, generating more heat, or even making audible noises. Like any malfunctioning mechanical device.
SSDs instead, with no mechanical parts, often just die; which makes SMART less effective. There are some signs (reallocation count, number of power failures, CRC failure checks), but it's nowhere close to the amount of signs an HDD can give.

Exactly. Platter based hard disks often have a five year warranty now. Do ssds have the same?

Samsung SSD warranty, up to 10 years: http://www.samsung.com/global/business/semiconductor/minisite/SSD/global/html/support/warranty.html

Intel SSD warranty, 3 or 5 years: http://www.intel.com/support/ssdc/hpssd/sb/CS-029645.htm

Crucial SSD warranty, 3 years: http://www.crucial.com/usa/en/company-warranty