HDD pauses at every command

[AusWolf]

This issue is a bit strange.

I noticed that one of my HDDs freezes for 10-20 seconds after I issue a command (opening a folder/file, initiating a copy, etc.) while showing 100% usage and 0 MB/s transfer rate. Sometimes even the file explorer stops responding until the HDD comes to its senses. After that, everything is fine. It also happens when watching a video that is on that disk. The video plays fine for a while, then when the player issues a buffer command, the video freezes, HDD usage is at 100% for a while, then the HDD comes back to life and works again.

There are no bad sectors, or other errors showing in SMART, and there's no heat issue as the drive sits in the HDD cage right in front of my chassis intake fans.

All drive power saving features are disabled, and HDD turn-off time is set to zero (disabled) in Windows.

Is this some kind of software bug that only plagues that one drive for some reason, or is the HDD on its way out?

 

[Solaris17]

Likely on the way out TBH. SMART isnt instantaneous and controller failures wont even register since support for controller health is so obscure or not even offered by the drive. Most SMART attributes exposed are for media only, which again isnt instant. You can go from 0 to 600 bad sectors for example on an HDD "in an instant" after it polls, but might have been struggling with the system for awhile. (its done generally on an "X amount of online hours cadence" so to the drive that may be 30hours but human perceivable is like 2 weeks worth of using a machine sporadically.)

You can try backing up (which you should do anyway) and formatting it to rule out any kind of odd corruption.

 

[AusWolf]

Likely on the way out TBH. SMART isnt instantaneous and controller failures wont even register since support for controller health is so obscure or not even offered by the drive. Most SMART attributes exposed are for media only, which again isnt instant. You can go from 0 to 600 bad sectors for example on an HDD "in an instant" after it polls, but might have been struggling with the system for awhile. (its done generally on an "X amount of online hours cadence" so to the drive that may be 30hours but human perceivable is like 2 weeks worth of using a machine sporadically.)

You can try backing up (which you should do anyway) and formatting it to rule out any kind of odd corruption.

I'll do that, thanks for the help.

It would be a shame for the drive to die so young - it's only about 7-8 months old, although, I've had worse.

I've always thought that HDDs are more reliable than SSDs for data storage, but I'm starting to doubt it.

 

[Legacy Ivy Terascale]

have you tried Crystal Disk Info ?
have you tried defragmenting ?
is it 90+ % full ?

 

[AusWolf]

have you tried Crystal Disk Info ?

Yes, all looks good there.

have you tried defragmenting ?

Yes, no change.

is it 90+ % full ?

No, it's around 50-60% full.

 

[Legacy Ivy Terascale]

can you put some photos with the computer , with the side panels removed , from multiple angles , including around the HDD area ?

 

[LabRat 891]

Disable power savings and drive power down.
Check/replace SATA cable.
Cleanly Reinstall SATA driver (if applicable).

Failing that fixing the issue, RMA/replace the drive.

Twice, I've had a drive pass all testing and have no cable issues, but be 'stuck' in some sort of ultra-low performance mode.
At least Windows was not reporting PIO mode, a common issue that can pop up w/ failing (even older) HDDs.

 

[JWNoctis]

The only time I've seen something similar, it was a failing laptop drive which was replaced under warranty. SMART showed no clue either that time, and the symptom is close: freezes at random, 100% activity, 0 data transfer, recover on restart, and if lucky.

Twice, I've had a drive pass all testing and have no cable issues, but be 'stuck' in some sort of ultra-low performance mode.
At least Windows was not reporting PIO mode, a common issue that can pop up w/ failing (even older) HDDs.

Is PIO mode still possible under SATA/AHCI at all, or is it unrelated? The only time I've seen that, it was a loose (or maybe pinched) PATA/IDE cable to an optical drive.

 

[LabRat 891]

Is PIO mode still possible under SATA/AHCI at all, or is it unrelated? The only time I've seen that, it was a loose (or maybe pinched) PATA/IDE cable to an optical drive.

Yes/No.
I believe a SATA HDD connected to a motherboard w/ a Legacy BIOS in IDE/PATA compatibility mode, can still go into PIO mode.
-and I've seen way too many SATA-native OEM PCs "from the factory" in IDE/PATA mode.
A common 'fix' for major performance issues was (properly) setting up the customer's BIOS and Windows Vista/7 Install to utilize AHCI, instead.

 

[AusWolf]

Disable power savings and drive power down.
Check/replace SATA cable.
Cleanly Reinstall SATA driver (if applicable).

Failing that fixing the issue, RMA/replace the drive.

Twice, I've had a drive pass all testing and have no cable issues, but be 'stuck' in some sort of ultra-low performance mode.
At least Windows was not reporting PIO mode, a common issue that can pop up w/ failing (even older) HDDs.

Power down is disabled already.

I'll try with a different cable, see if it does anything - I have my doubts considering that the drive works fine when it is actually doing something. It's only when I give it a command after some idle time when the problem starts.

I'll check for SATA and chipset drivers, see if an update works.

Yes/No.
I believe a SATA HDD connected to a motherboard w/ a Legacy BIOS in IDE/PATA compatibility mode, can still go into PIO mode.
-and I've seen way too many SATA-native OEM PCs "from the factory" in IDE/PATA mode.
A common 'fix' for major performance issues was (properly) setting up the customer's BIOS and Windows Vista/7 Install to utilize AHCI, instead.

There's definitely no IDE compatibility mode on this system. It's not OEM, I always build everything myself.

Besides, I've got two HDDs, and the second one works fine. IDE mode would affect both of them, wouldn't it?

 

[basco]

do ya have something like this disabled in bios?
Aggressive SATA Sleep on AMD, or Aggressive Link Power Management for SATA controller on Intel

 

[Calenhad]

One other thing you could try is boot from a Linux usb live media and see if you can replicate the issue there.

Otherwise I agree with others, if a new cable does not fix this, that drive is probably on it's way out. My guesstimate is the controller, not the physical media, is going bad.

 

[Tahagomizer]

What immediately comes to mind is the disk insists on entering power saving mode (idle spindown) which, from experience, often means a problem with heads, notably head opamps. It doesn't manifest itself in S.M.A.R.T until the controller can't see the heads any more and the drive is just dead.
It also might be an interface problem, damaged cable, faulty power delivery, bad mojo in the room. Check with another drive, check the drive in another machine, above all don't trust this one - you do, of course, have a backup, right?

 

[Japie073]

I had few drives set up in RAID0 on one of my previous builds on a Z97 board that had the same issue. SATA link state power management causes one of the drives to spin up and then down causing lag. Check under power settings in Windows check that the setting to turn off hard disks are set to 0.

 

[Legacy Ivy Terascale]

This issue is a bit strange.

I noticed that one of my HDDs freezes for 10-20 seconds after I issue a command (opening a folder/file, initiating a copy, etc.) while showing 100% usage and 0 MB/s transfer rate. Sometimes even the file explorer stops responding until the HDD comes to its senses. After that, everything is fine. It also happens when watching a video that is on that disk. The video plays fine for a while, then when the player issues a buffer command, the video freezes, HDD usage is at 100% for a while, then the HDD comes back to life and works again.

There are no bad sectors, or other errors showing in SMART, and there's no heat issue as the drive sits in the HDD cage right in front of my chassis intake fans.

All drive power saving features are disabled, and HDD turn-off time is set to zero (disabled) in Windows.

Is this some kind of software bug that only plagues that one drive for some reason, or is the HDD on its way out?

do you have them in RAID ( 0 1 5 10 ... ) ?

 

[AusWolf]

I've just reinstalled the AMD chipset driver. Everything seems to be working fine for now. I just don't understand how the chipset driver can affect one HDD, but not the other (I have two), so I'll keep an eye out.

One other thing you could try is boot from a Linux usb live media and see if you can replicate the issue there.

Otherwise I agree with others, if a new cable does not fix this, that drive is probably on it's way out. My guesstimate is the controller, not the physical media, is going bad.

I've been wanting to practice with Linux anyway, so when I do, I'll make sure to test it.

do you have them in RAID ( 0 1 5 10 ... ) ?

No, just simple, individual SATA connections.

 

[remixedcat]

is it the drive in your profile specs?

 

[AusWolf]

is it the drive in your profile specs?

Yes, the 4 TB one. It's about 7-8 months old.

 

[Hardcore Games]

SATA SSD prices are so low that I removed low capacity hard disks in favor of solid state drives which run faster as well

4TB SATA SSD are $249 in Canada and prices are falling fast

 

[chrcoluk]

The symptom indicates either drive having to spin up or unparking the head, head unparking is quicker than spin up though maybe 1-3 seconds.

What does SMART report for load/unload cycle count?

Note head parking on Seagate and WD (quite possibly other hdd's also), is independent of OS power saving, also independent of SATA ALPM. Both vendors supply a tool to change its behaviour.

If it is a faulty drive, the fault rate is actually at its lowest during mid life, so you more likely to get a faulty HDD at under 1 year old compared to say 3-5 years old.