HDD SMART variable was FAILING but is now OK after writing 0s to it

[Alir]

I have made a recent thread about a failing HDD. That one is different to this one.

 

One of my internal HDDs was emitting the sounds of hell a few days ago. It was clicking a lot for hours whilst I was copying files off, as if the read head was hitting the platter - or the read-write head was constantly retreating away from the platter. It was a clicking sound basically and didn't sound pleasant at all.

 

I was getting notifications in Linux that my drive was failing and one SMART attribute was failing. I believe it was the "Seek Error Rate". I used my expert Googling skills and found that it should not affect the integrity of my files but did take ages to copy files over to my backup HDD. I bought a spare HDD in a hurry so that I can copy the files onto there. Once I copied the files off, I was expecting to have the drive recycled. I wrote 0s to the entire drive with dd and rebooted Linux (a Live CD), the clicking has completely gone. SMART tests say my drive is "OK" now and the SMART attribute itself is "OK". 

 

How the heck did that happen? The SMART attribute before was "FAILING" and Linux was warning me my drive was about to fail. I have never known SMART variables reset for any reason. 

[RobbinM]

It might be a hardware thing. Completely overwriting it with zeroes means the head has travelled over all the surface of the platters, reached it outer limits. 

I would still not trust the drive enymore because it might be a temporary fix. But it can be a fun experiment to see how long it will last before it finally breaks. Maybe put it in a RAID1 for fun.

[Alir]

On 9/11/2018 at 2:44 AM, RobbinM said:

It might be a hardware thing. Completely overwriting it with zeroes means the head has travelled over all the surface of the platters, reached it outer limits. 

I would still not trust the drive enymore because it might be a temporary fix. But it can be a fun experiment to see how long it will last before it finally breaks. Maybe put it in a RAID1 for fun.

That didn't really answer the question as to why this would happen.

 

Thing is I don't trust the drive for anything really. 

[RobbinM]

Is your answer in here possibly?

[Alir]

On 9/12/2018 at 12:26 PM, RobbinM said:

Is your answer in here possibly?

 

I think so

[STRESSMASTER]

Spoiler

On 9/10/2018 at 9:29 PM, Alir said:

I have made a recent thread about a failing HDD. That one is different to this one.

 

One of my internal HDDs was emitting the sounds of hell a few days ago. It was clicking a lot for hours whilst I was copying files off, as if the read head was hitting the platter - or the read-write head was constantly retreating away from the platter. It was a clicking sound basically and didn't sound pleasant at all.

 

I was getting notifications in Linux that my drive was failing and one SMART attribute was failing. I believe it was the "Seek Error Rate". I used my expert Googling skills and found that it should not affect the integrity of my files but did take ages to copy files over to my backup HDD. I bought a spare HDD in a hurry so that I can copy the files onto there. Once I copied the files off, I was expecting to have the drive recycled. I wrote 0s to the entire drive with dd and rebooted Linux (a Live CD), the clicking has completely gone. SMART tests say my drive is "OK" now and the SMART attribute itself is "OK". 

 

How the heck did that happen? The SMART attribute before was "FAILING" and Linux was warning me my drive was about to fail. I have never known SMART variables reset for any reason. 

 

This is usually caused by a head getting stuck. This sometimes reduces the HDD spin speed. (thus low performance) By writing every bit on the HDD what probably happened is that you were able to get the head to un-sick thus it fully works (for now) It has been my experience that the drive will fail within a year usually as the head most likely got damaged (50/50 chance).

[Dionysos]

Don't even try to revive an HDD once it starts running slowly (same thing to SSD), logging critical errors, or making weird sound. Its SMART log might not show honestly (vendor fake or hide). Backing up or replacing it ASAP is the best policy.

[Alir]

14 hours ago, STRESSMASTER said:

  Reveal hidden contents

 

This is usually caused by a head getting stuck. This sometimes reduces the HDD spin speed. (thus low performance) By writing every bit on the HDD what probably happened is that you were able to get the head to un-sick thus it fully works (for now) It has been my experience that the drive will fail within a year usually as the head most likely got damaged (50/50 chance).

Would that explain the slow read writes now? To clarify, the HDD is not making sounds anymore and the SMART variables are all fine. but it is very slow.

 

7 hours ago, Dionysos said:

Don't even try to revive an HDD once it starts running slowly (same thing to SSD), logging critical errors, or making weird sound. Its SMART log might not show honestly (vendor fake or hide). Backing up or replacing it ASAP is the best policy.

Got it. Will probably leave it in the PC for random Linux installs when I want to try out a Linux distro but am not actually planning on using it for anything