Unsafe Shutdown Count...

[stahlhart]

We had a power failure late last night, and sure enough, when I powered the workstation up again this morning after it was restored, the counter in SMART ticked up one again. I'd been thinking about adding a UPS for some time now, and this last outage pretty much made the decision for me.

But the SSD doesn't seem to be any worse for wear otherwise -- speed test results are about the same, and SSDLife still hasn't budged from 99%. Is there any empirical data for how high an unsafe shutdown count one can have before the drive is at risk for damage? Does it even matter for most users? Why does SMART track this?

 

[Spikesoldier]

probably a carryover from the time before ssds

ive been eyeing a UPS for my fileserver. freenas doesnt like power outages.

 

[Coup27]

Which SSD?

It will not damage your SSD physically in anyway having the power removed out of the blue. Depending on the SSD, the worst that will happen is your may lose some data which has not been wrote to NAND yet. You're actually going to wear the SSD out quicker by panicing and benchmarking it.

Although adding a UPS is hardly a bad thing. You can get them really small now and quite cheap.

Something you might be interested in, I recently sold my X25-M G2 but I used that for about 3 years and only had a power outage once, yet unsafe shutdown count was in double figures. Not sure how accurate or important the figure is.

 

[corkyg]

SMART doesn't really bring much useful info to the party where SSDs are concerned. And, SMART implementation is not really standard throughout the industry. Bottom line - a UPS is a good investment.

 

[stahlhart]

Which SSD?

It will not damage your SSD physically in anyway having the power removed out of the blue. Depending on the SSD, the worst that will happen is your may lose some data which has not been wrote to NAND yet. You're actually going to wear the SSD out quicker by panicing and benchmarking it.

Although adding a UPS is hardly a bad thing. You can get them really small now and quite cheap.

Something you might be interested in, I recently sold my X25-M G2 but I used that for about 3 years and only had a power outage once, yet unsafe shutdown count was in double figures. Not sure how accurate or important the figure is.

I also have an X-25-M G2, 120Gb. My count is in double figures also; most of them probably BSODs from overclocking experiments.

More concern than panic, mainly because the power supply in my wife's desktop gave up the ghost under similar circumstances about a year and a half ago.

 

[Mark R]

Intel SSDs, for some reason, have an unsafe shutdown counter. This isn't something that other manufacturers have on either HDD or SDD.

All this counter shows is how many times the SSD has lost power, without having been sent a "shutdown" command from the OS. It isn't measuring any harm to the drive, but can be a marker of an unstable system. An OS will typically send a "shutdown" command to a hard drive, to guarantee that any data in the drive's cache has been saved, before the OS allows the power to be turned off. Without this command, data can occasionally get lost (because it never actually got saved onto the platters/flash, even though the OS thought it had been saved).

My Intel has an "unsafe shutdown count" in the hundreds, because a glitchy graphics card, would cause the BIOS to hang at POST. Every time I powered off the comp because it got stuck at the BIOS, the "unsafe shutdown count" went up by 1.

 

[stahlhart]

Intel SSDs, for some reason, have an unsafe shutdown counter. This isn't something that other manufacturers have on either HDD or SDD.

All this counter shows is how many times the SSD has lost power, without having been sent a "shutdown" command from the OS. It isn't measuring any harm to the drive, but can be a marker of an unstable system. An OS will typically send a "shutdown" command to a hard drive, to guarantee that any data in the drive's cache has been saved, before the OS allows the power to be turned off. Without this command, data can occasionally get lost (because it never actually got saved onto the platters/flash, even though the OS thought it had been saved).

My Intel has an "unsafe shutdown count" in the hundreds, because a glitchy graphics card, would cause the BIOS to hang at POST. Every time I powered off the comp because it got stuck at the BIOS, the "unsafe shutdown count" went up by 1.

Okay, so it's all in the perception of where the risk lies, and in this case it's data and not hardware. That's reassuring.

Thanks, everyone, for the feedback...