Has my HC2 become a drive killer?

[DouglasQuaid2]

After a couple of years of smooth operation, all my dockers started having issues after a period of time after reboot.
I found this was because the 4TB WD RE that hosted all the files and config was going read only due to drive errors, and the S.M.A.R.T status was showing a raw reallocated sectors slowly rising.

This did seem a bit strange, as these drives are very hard to kill. I have 6 more of them in other devices and they are all still at %100 and going strong.

Ultimately deciding I must have just been unlucky with this one, I cloned the drive to a brand new 4TB WD Red Plus NAS drive, straight out of the packet, and through the WD RE in the trash.

To my dismay, it seems the same thing is now happening.
Dockers stopping working due to file system going read only, and S.M.A.R.T is showing 'Current Pending Sector' from 0 to 5, now 10.

The probability of having a flaky WD RE, followed by a faulty brand new WD Red Pro NAS is extremely low.
Note. The drive is secured firmly to the big heatsink, I even applied a small bit of compound this time to make sure. And I've left the cover off. The drive temp fluctuates between about 32c to 38c when streaming video from it so I dont think its a temperature thing.
WD Red Plus specs say designed for constant higher temps (though I can't see any numbers)

Has something happened to the HC2 that's made it a drive killer?!


Only hours old!

[mctom]

Hi!

What you reported looks like a physical layer fault of a mechanical drive. There isn't much that HC2 can do to kill it this way. The only thing HC2 could do is either provide wrong supply or data line voltages, none of which could possibly cause physical errors on the drive plates.
Temperature shouldn't be an issue unless the drive goes so hot you can't hold it in your hand for long (which is around 50C for metal objects).

Perhaps there is something that has changed in the vicinity of your setup, such as a strong magnetic field, or something vibrating.

Have you tried running a "faulty" drive on another machine, to see if it continues deteriorating?

[DouglasQuaid2]

Hi!

What you reported looks like a physical layer fault of a mechanical drive. There isn't much that HC2 can do to kill it this way. The only thing HC2 could do is either provide wrong supply or data line voltages, none of which could possibly cause physical errors on the drive plates.
Temperature shouldn't be an issue unless the drive goes so hot you can't hold it in your hand for long (which is around 50C for metal objects).

Perhaps there is something that has changed in the vicinity of your setup, such as a strong magnetic field, or something vibrating.

Have you tried running a "faulty" drive on another machine, to see if it continues deteriorating?

Yes what you are saying were my initial thoughts, and exactly the reason for buying a brand new WD Red Plus 4TB NAS drive to replace the one that presumably had "physical layer fault of a mechanical drive", copying the files across, and throwing the original 'faulty' drive in the trash.

The chances of a WD RE enterprise drive going bad are slim, but it happens. It would indicate a fault/degradation with age, of the drive, not the Odroid.
The changes of an WD RE enterprise drive going bad, then replacing it with a brand new WD Red Plus 4TB NAS, and it going instantly bad in exactly the same way that the last one did are extremely improbable and would indicate a fault with the device they were plugged in to.
I'm even regretting discarding the other drive now, as it probably would still work fine in another unit

I guess the first thing to replace would be the wall wart (that I ordered along with the device from Odroid)

[mctom]

Too bad I've never had a broken HDD in my entire life and don't know much about its fault modes. My comment is solely based on my understanding of the principles of HDD operation.
And like I said earlier, I don't think it has anything to do with voltage levels or such, especially in enterprise HDDs that are surely hardened against transient out-of-spec events.

Is there any possibility that issuing wrong commands to the HDD can damage it? I know it was a thing in the DOS era, when a computer virus could damage a HDD, but these days, and without a specific intention to do so? Also super unlikely.

Out of curiosity, does your new HDD bear a manufacture date, and is in fact newer than your broken one?
I googled WD Red Plus 4TB and read a bunch of instant fault reports on them. Some users recommend not getting them, possibly because they've acquired mistrust to the brand in the past. Some suspect planned obsolescence.