SMART reporting and Hard Disk buzzing sound

  • Thread starter Thread starter Vanadium 50
  • Start date Start date
  • Tags Tags
    Disk Hard Sound
Click For Summary
A Toshiba DT01ACA300 hard drive is exhibiting concerning signs of failure, including frequent buzzing sounds and an alarming SMART report indicating 61 million read errors. Despite the SMART health assessment passing, the drive has shown a Raw_Read_Error_Rate that has fluctuated, and 87 sectors have been reallocated recently. The drive's performance has degraded, taking significantly longer to read data than usual, which raises red flags about its reliability. Users suggest that while some read errors are normal, the sheer volume of errors and the drive's increasing noise level indicate a potential imminent failure. Recommendations include backing up data immediately, replacing the drive, and avoiding frequent RAID verification checks that may stress consumer-grade drives. The discussion highlights the importance of monitoring drive health and being proactive about data protection, especially with aging hardware. Additionally, there are concerns about the reliability of SSDs and the practices of drive manufacturers, particularly regarding the quality of components used in consumer-grade products.
  • #31
fresh_42 said:
Thanks. Looks like an internal. They seem to be far less reliable than external ones (my experiences).
An external USB HDD these days usually contains a SATA internal HDD physically. I have a Seagate 4TB drive that used to be in an enclosure. The exact same drive was available at a higher price without the enclosure. Perhaps the USB SATA interface in the enclosure is less demanding than a direct internal SATA connection might be, but that is entirely an interface concern, and has nothing to do with the drive itself.
 
Computer science news on Phys.org
  • #32
Vanadium 50 said:
Another update.

The raw read error rate is not the raw read error rate. :doh: It's actually logarithmic. A value of 73 means that you get an error every 20,000,000 reads. (10^7.3). Failing at 006 means every fourth read fails, and I would not use the word "failing" to describe that. More like "failed".

I just replaced a 320 GB Seagate DB35.3 that was in 24/7 use for 12 years. The only problem was that SMART tests took a (very) long time to finish, but after 12 years I figured I got my money's worth. The replacement isn't exactly new - it is a Seagate 500 GB that Amazon listed as New, but it wasn't new old stock or even new: it had 100 hours on it. And a filesystem. Amazon refunded my money and told me not to bother shipping it back. Can't complain about the service, but I suspect the business model is to sell lightly used drives as new, and do a refund only if someone notices and complains.
Amazon was probably merely passing forward the supplier's assessment of the supplier's product. if you want to critique Amazon regarding the matter, you might consider that the supplier was apparently not adequately pre-vetted, and that refunds aren't always an adequate remedy. I think that Amazon could do better.
 
  • #33
sysprog said:
Amazon was probably merely passing forward the supplier's assessment of the supplier's product.

I agree with that, except for the word "merely". That's Amazon's job.
 
  • #34
sysprog said:
The exact same drive was available at a higher price without the enclosure.

The4re is a whole counter-culture of people who "shuck" drives- buy a USB drive and remove the drive. When I replaced my years-old external for a larger one, I disassembled the unit to take a look inside. The USB-to-SATA connection is tiny, and if it didn't have an LED on it you could be excused for missing it.

The drive vendors have caught on to this, and their moral equivalent of the airlines' "Saturday night stay" is to disable drives if the 3.3V line is present. This is fairly simple to work around, although I have to question whether using a $30 roll of Kapton tape is really the best solution to save $15 on a drive.
 
Last edited:
  • Wow
Likes sysprog
  • #35
Vanadium 50 said:
The4re is a whole counter-culture of people who "shuck" drives- buy a USB drive and remove the drive. When I replaced my years-old external for a larger one, I disassembled the unit to take a look inside. The USB-to-SATA connection is tiny, and if it didn't have an LED on it you could be excused for missing it.

The drive vendors have caught on to this, and their moral equivalent of the aorline's "Saturday night stay" is to disable drives if the 3.3V line is present. This is fairly simple to work around, although I have to question whether using a $30 roll of Kapton tape is really the best solution to save $15 on a drive.
On seeing this from you, I looked it up, and found that you can bypass the 3.3V rail by using a 4-pin Molex-to-SATA adapter to supply power to the HDD.
 
Last edited:
  • #36
Yes, although modern supplies tend to have few Molex connectors. A SATA-to-SATA extension without an orange wire (or with a cut orange wire) will do the same thing.
 
  • Like
Likes sysprog
  • #37
Vanadium 50 said:
Yes, although modern supplies tend to have few Molex connectors. A SATA-to-SATA extension without an orange wire (or with a cut orange wire) will do the same thing.
The converter cables/adapters are not too expensive -- this one from Monoprice is $1.29 (USD):

1581280089033.png


I understand that the HDD might require the SATA end and the power supply might not like the MOLEX end, but one way or another, you can still use the HDD outside of the enclosure.
 
Last edited:
  • #38
I don't know the origin of this but here it is:

##\mathtt{\text{If you have a problem and it can't be fixed with duct tape,}}##
##\mathtt{\text{it's 'cause you're not usin' enough duct tape.}}##​
 
  • Like
Likes russ_watters
  • #39
One thing to keep in mind is that there are a lot of junk adapters out there that use wire that is too small. Most of them won't catch fire...

SATA is rated at 4.5 amps. Molex is rated at 11 amps. AWG23 would be the minimum gauge wire to power one SATA within spec. Most drives are < 10 W, so some manufacturers use AWG 26 ("probably" safe) or even AWG 28 (hey, there's a safety margin of 2, right?).
 
  • Informative
Likes sysprog
  • #40
The twin of the RMA'ed drive back in July is throwing read errors. Grrrr...
 
  • Wow
Likes russ_watters