Intel 330 SSD "Unsafe Shutdown Count" very high.

[golem]

Although this SSD remains flawless in nearly 2 years of operation I did, while performing my quarterly "Intel SSD Toolbox" tune, notice the Unsafe Shutdown Count is abnormally high. There is absolutely no outright indication of non-systematic shutdown. I can add that this drive is partitioned for dual booting. Running this diagnostics from each bootable partition provides nearly identical results.

SSD Toolbox along with all hardware and device drivers are regularly checked to be up to date. Multiple times a day I do mount and dismount external USB HDD's and also SATA HDD's via hot-swap bays but can't envision this being a factor in the high count. Unfortunately as I just noticed this high value it's impossible to determine if it was installed with a high count or gradually accumulated. As it nearly mirrors the power cycles I assume it was gradual accumulation.

Any insight, simply out of curiosity, would be much appreciated.

 

[ignatzatsonic]

I have an Intel 320, a little over 3 years old.

Unsafe shutdowns shown as 54. Frankly, I hadn't paid attention to it before, so I don't know how that number has moved over time.

My data drive, a Samsung 103SJ spinner, shows NO unsafe shutdowns and it's been installed longer than the Intel. Any "unsafe shutdown" on the Intel should affect the Samsung equally, yet it's 54 to 0.

My backup drive is only 3 weeks old and shows 1 unsafe shutdown. I can't explain that either.

I have a very vanilla system--3 internal drives, each with a single partition. Every month or two, I'll connect an external dock and run some backups to another HD.

I have a power outage maybe twice a year while the PC is running.

All I can suggest is to bring it to Intel support's attention by phone or email and see what they say. I'd be interested myself. My best guess is that the number is either erroneous or irrelevant.

 

[golem]

Much appreciate the reply ignatzatsonic!

Not sure why I didn't but you got me thinking to check the two internal HDD's for comparison. While they are a fair bit older than the SSD they do exhibit counts to the extremes. The first is an Hitachi "CoolSpin" with 8,471 and the second is a WD "Green" (EARX) displaying a svelt 25 count for the 0xC0 Hex.

This really may not be an apples to apples comparison in regards to the SSD vs HDD when it comes to this particular S.M.A.R.T test. The attribute is listed as -- "Power-off Retract Count, Emergency Retract Cycle Count (Fujitsu), or Unsafe Shutdown Count - Count of times the heads are loaded off the media. Heads can be unloaded without actually powering off." Not sure whatthe comparability is. Interestingly the WD "Green" drives were maligned for the head parking issue but for me that drive seems to be the least offender (maybe it was the EARS drives).Not sure I'm going to pursue this as something requiring attention unless someone pipes in and is adamant that this is a detrimental situation. Just for the heck of it I attached some screen shots of S.M.A.R.T results using karlsnooks' (from this forum) Speccy app.

Order listed below are:
Intel 330 SSD
Hitachi HDS5C3020ALA632
WDC WD20EARX-008FB0

-

 

[ignatzatsonic]

Hmmm.........."power off retract count".

I'd never heard of it.

But I downloaded Speccy and it seems that what the Intel Toolbox SMART details show as "unsafe shutdown count" is called "power off retract count" in Speccy--for both my Samsung and WD Green HDDs.

Both Speccy and Intel Toolbox use the term "unsafe shutdown count" for my Intel SSD.

In all cases the attribute is IDed as "C0", so I assume the 2 terms are effectively synonyms.

Yet another instance of SMART being unnecessarily arcane and confusing.

I know there are other Intel SSD owners perusing this forum, so maybe you'll get more comments.

 

[golem]

Hmmm.........."power off retract count".

I'd never heard of it.

But I downloaded Speccy and it seems that what the Intel Toolbox SMART details show as "unsafe shutdown count" is called "power off retract count" in Speccy--for both my Samsung and WD Green HDDs.

Both Speccy and Intel Toolbox use the term "unsafe shutdown count" for my Intel SSD.

In all cases the attribute is IDed as "C0", so I assume the 2 terms are effectively synonyms.

Yet another instance of SMART being unnecessarily arcane and confusing.

I know there are other Intel SSD owners perusing this forum, so maybe you'll get more comments.

Ha! I was going to snip your quote but it seems nearly all of it is pertinent to my following comment -- I can actually see where the two differing attribute definitions would be applicable to the corresponding drive types but do find it difficult to grasp how (or why) SMART would not have given each of these an independent HEX/ID designation. I can easily envision "Power-off Retract Count" and "Unsafe Shutdown Count" as completely different actions. In the end I'm sure it's wholly possible to double-dip a HEX/ID and use it for two purposes but as you say it can lead to ambiguous and/or confusing interpretation.

Thanks again for your reply. I'll continue research on this, but not fervently.

*EDIT* -- Just got to thinking about my comment on the low count of 25 for WD and not equating with expected head parking count. I was probably overlooking the fact that the "Load/Unload Cycle Count" is likely what would reflect this, and it sure looks to. For some it probably wouldn't take too long to reach the 300k life-cycle rating.

 

[ignatzatsonic]

Golem:

Looking over your pictures, I have some questions.

First pic shows "power on hours" for the Intel SSD as over 900,000; which is circa 102 years for a drive that was manufactured 2 years ago. What's the proper interpretation of that apparent error?

The Hitachi pic shows identical value of 8471 for Power Off Retract and Load/Unload Cycle. Is that mere coincidence or are the two related? I assume it's coincidence since the WD Green shows major differences for those two attributes.

I had assumed that LLC was related to head parking, but I'm not clear on what Power Off Retract allegedly measures. "Retract" referring to heads retracting? "Power Off" referring to shutdown of some type?

Could SMART possibly be more hopeless for a layman?

 

[golem]

ignatzatsonic, seems you are like me..Glutton for punishment.

First pic shows "power on hours" for the Intel SSD as over 900,000; which is circa 102 years for a drive that was manufactured 2 years ago. What's the proper interpretation of that apparent error?

Quoting from the S.M.A.R.T. Wiki -
"The raw value of this attribute shows total count of hours (or minutes, or seconds, depending on manufacturer) in power-on state."
Based on that I'm assuming the count is in seconds which would work ot to roughly 250h. Still not wholly in concert with Speccy but close.

The Hitachi pic shows identical value of 8471 for Power Off Retract and Load/Unload Cycle. Is that mere coincidence or are the two related? I assume it's coincidence since the WD Green shows major differences for those two attributes.

It seems both are related to the heads being retracted to the landing position. "Power-off Retract" landing should occur each time the power is turned off while with "Load/Unload Cycle Count" landing can also occur whenever the manufacturer has designed it to do so during an idle state (could be as short as 5sec). That could explain why the count is similar for the Hitachi but then gives pause as to why the WD is so low as I power this system down almost every evening (actually twice a day if booting into the other partition).

I had assumed that LLC was related to head parking, but I'm not clear on what Power Off Retract allegedly measures. "Retract" referring to heads retracting? "Power Off" referring to shutdown of some type?

See reply above. [Again this is me "assuming" my understanding is correct].

Could SMART possibly be more hopeless for a layman?

I think not. Definitely looks not to be a black-n-white what-you-see-is-what-you-get ease of interpretation.

 

[ignatzatsonic]

To add more weirdness:

My Intel 320, older than your 330, shows the power on hours as 16513 in the Toolbox and as 688 days, 1 hour in Speccy. This is about right. I run it maybe 14 or 15 hours a day and I've owned it a little over 3 years.

What's weird is that Intel was sober enough to correctly show hours in an attribute labeled "power on hours count" when my 320 was made (2011), but seems to have been drunk enough to show the count in seconds (possibly) for an attribute labeled as "power on hours count" on your 330 a year or so later.

Regarding load cycles, unsafe shutdown count, and power off retract count on WD Greens. Below is what I show for the 2 programs on my 3 week old Green 3 TB. Make of it what you will.

I've got my good eye trained on the load/unload cycle count for this drive. As you know, there was that issue on some Greens of head parking too often. And there is that tool to set the default to a higher value in seconds to prevent too much parking. I've got the tool downloaded but have not yet used it on this new drive----because I haven't seen anything definitive on whether or not current generation Greens have the "too much head parking" issue. Mine seems under control so far--a few hundred in a few weeks, certainly well within the 300k rating.

But I've read of cases where earlier Greens were seemingly OK on load/unload cycle count and then went berserk for no apparent reason, running up many thousands of cycles per month. So...........