2

I have a secondary hard drive in my Windows 10 machine which has not been given any problems until it suddenly disconnected today. After a shutdown and restart, everything appears to be working and the only error/warnings messages in the event log were

An error was detected on device \Device\Harddisk0\DR0 during a paging operation.

Followed by

Disk 0 has been surprise removed.

(Note , the Disk is Harddisk0\DR0, but is not the bootable/system disk with is a separate SSD drive)

I tried to run a few tests to see if the disk is bad or is about to go bad and the results don't seem to be consistent.

Chkdsk said there was no need to check it as no errors had been detected, but I continued and it completed without errors.

wmic /namespace:\\root\wmi path MSStorageDriver_FailurePredictStatus

returns false.

DiskSmartView (from Nirsoft) seems to say there are no re-allocated sectors, although it does have a Unknown status rather than OK.

DiskSmartView

CrystalDiskInfo says the Health Status is good, but appears to suggest the number of re-allocated sectors is 200 which I would not consider good.

CrystalDiskInfo

I must be misreading one of these programs, but I am not sure which.

So the question is - Is my disk on its way out?

Added SmartCtl results SmartCtl

sgmoore
  • 6,423
  • 2
  • 25
  • 33
  • Use `smartctl` directly, as CrystalDisk doesn't accurately show values _(HDD temp is not 123C)_. The main values to pay attention to for drive health are `Uncorrectable Sector Count`, wear level if SSD, Hardware errors section, self-test log, and Power-On Hours. Download [`smartmontools`](https://www.smartmontools.org/wiki/Download#InstalltheWindowspackage) → `smartctl -a /dev/` → Run short test: `smartctl -t short /dev/` → Get disk status after 120s: `smartctl -a /dev/` → Run long test _(takes several hours)_: `smartctl -t long /dev/` → Re-check disk info once finished – JW0914 Mar 10 '22 at 17:23
  • _(Cont'd...)_ There's no issue with the drive. If `Reallocated Sector Count` had a RAW value, it simply means a read/write failure occurred, not that it occurred due to hardware [drive] damage _(bad cable, transient issue with cable shielding resulting in data corruption, connector seating, etc. can all cause read/write errors)_; while this would indicate a data issue somewhere, without uncorrectable sectors or hardware errors, it's likely not drive-related - when physical [hardware] damage occurs, it's seen in the uncorrectable sector count, hardware errors section, and self-test log section. – JW0914 Mar 10 '22 at 17:31
  • 1
    For intermittent connectivity, always first suspect the SATA cable or socket. They're a terrible design & don't like being poked, prodded or bent. – Tetsujin Mar 10 '22 at 17:42
  • @JW0914 . Will run the long test shortly, but the short test gives exactly the same results as CrystalDisk including the 123 Temperature in the Value column, but the actual temperature 27 is in the raw_value column. (The advantage of smartctl is that it defaults to show the raw values in decimal, whereas CrystalDisk defaults to hex, but there is an option to display in decimal) – sgmoore Mar 10 '22 at 17:43
  • @sgmoore That's abnormal, as the [Current] _Value_ should be the current temp in C and _Worst_ should be the highest temp in C it's reached during its current power cycle. I'm simply not a fan of CrystalDiskInfo because it modifies the format of S.M.A.R.T data when S.M.A.R.T data has a standardized presentation/output, repeatedly deviating from that standardization _(a user shouldn't need to read its help section to figure out what it's modified and what it hasn't - that's inefficient absurdity to me)_. `smartctl` is the main tool used for S.M.A.R.T data and it should mirror `smartctl` output – JW0914 Mar 10 '22 at 17:54
  • @JW0914 Not sure if your comment about CrystalDiskInfo is fair. I have added the smartctl results and I don't see much difference (apart from the hex/decimal issue). Tests were run at different times, so some differences would be expected. – sgmoore Mar 10 '22 at 17:57
  • @sgmoore _(Please don't screenshot terminal output, instead copy/paste it into a code box.)_ The point is there's a list of differences and there shouldn't be - the reason this is problematic: `smartctl` is the defacto tool used across all OSes [BSD, Linux, Mac, Windows] to poll S.MA.R.T data, run S.M.A.R.T tests, and perform HDD burn-ins, and when looking for help, 99.99% of folks who know how to read S.M.A.R.T values and use S.M.A.R.T info are going to be familiar with the standardized format of S.M.A.R.T data `smartctl` outputs, not the deviation CrystalDiskInfo does. – JW0914 Mar 10 '22 at 18:08

2 Answers2

3

The SMART data indicates that your disk is in perfect shape.

The algorithm calculates this number based upon the raw data, but it's not intuitive.

A new hard drive should have a high number which is the theoretical maximum (100, 200, or 253 depending on the manufacturer), that decreases during its lifetime.

The indicator for Temperature doesn't make sense, but apart from it everything is correct.

Take this event as a warning and keep backups, but the disk seems to be OK.

From NTFS.com S.M.A.R.T. Attributes:

Attribute values can range from 1 to 253 (1 representing the worst case and 253 representing the best). Depending on the manufacturer, a value of 100 or 200 will often be chosen as the "normal" value.

For most attributes, a value above the threshold is good and means no errors.

harrymc
  • 455,459
  • 31
  • 526
  • 924
  • Hmm. Two answers (from two people with high reps) and a big difference in opinion! Note sure if I understand you. Are you saying that the reallocated sectors count **decreases** as the disk during the drives lifetime. That seem to contradict the info on https://en.wikipedia.org/wiki/S.M.A.R.T. – sgmoore Mar 10 '22 at 16:46
  • I added a reference. Also from the Wikipedia link: "Each attribute has a raw value that can be a decimal or a hexadecimal value, whose meaning is entirely up to the drive manufacturer (but often corresponds to counts or a physical unit, such as degrees Celsius or seconds), a normalized value, which ranges from 1 to 253 (**with 1 representing the worst case and 253 representing the best**) and a worst value, which represents the lowest recorded normalized value. The initial default value of attributes is 100 but can vary between manufacturer". – harrymc Mar 10 '22 at 17:14
  • Never try to understand the RAW value yourself. Use a tool like Speccy that analyzes the value for you and tells if it's good or not. – harrymc Mar 10 '22 at 17:15
  • Thanks, I think I am being to understand now. – sgmoore Mar 10 '22 at 17:23
2

CrystalDiskInfo says the Health Status is good, but appears to suggest the number of re-allocated sectors is 200 which I would not consider good.

No, you are misreading the columns. The raw value in Crystaldiskinfo tells you that there are zero reallocated sectors.

If I read the report like you do the temperature of your drive is 123°C whereas in reality it is 1b(hex) = 16+11 = 27 °C. The temperature is perfect.

Furthermore your load/unload cyles (raw) value indicate that your use profile does not match the original parking time schedule of WD which is 8 seconds if I remember correctly.

WD has set a very short time period after which the firmware parks the heads. I can't find my documentation, I think it was 8 seconds - I own a WD40EZRZ myself. You drives experienced already ~240.000 cycles which is below the guaranteed figure of 300.000 (see link)

https://documents.westerndigital.com/content/dam/doc-library/en_us/assets/public/western-digital/product/internal-drives/wd-blue-hdd/data-sheet-wd-blue-pc-hard-drives-2879-771436.pdf

but you could avoided that situation by increasing the parking time interval. After only 3 years of running time you already reached 83% of the guaranteed load cycle figure.

To learn more about the issue search for WDIDLE3 or read

https://www.truenas.com/community/threads/hacking-wd-greens-and-reds-with-wdidle3-exe.18171/

As for the comment of JW014 below I prefer to use smartmontools over Crystaldiskinfo although you sometimes have to set additional parameters such as -d,sat or similar when the drive is connected by means of an external USB housing.

r2d3
  • 3,298
  • 1
  • 8
  • 24
  • Simply put, CrystalDiskInfo doesn't accurately display S.M.A.R.T info _(it shouldn't be modifying data garnished from `smartctl -a /dev/disk`, or modifying the format of the data, as S.M.A.R.T data has a standardized presentation/output and CrystalDiskInfo repeatedly deviates from that standardization)_, which is why I always recommend to use `smartctl` directly since [`smartmontools`](https://www.smartmontools.org/wiki/Download#InstalltheWindowspackage) is available for Windows. – JW0914 Mar 10 '22 at 17:38