Cause of SMART "Electrical Failure" on Seagate drives

[puterboy]

I bought several new Seagate ST31000340AS drives for use in a RAID array.
After a couple of days, I kept getting sudden Linux system crashes (despite the RAID array) and independent of whether I booted from the new boot array or an old system disk.

I then ran smartctl (SMART) test on the various disks.
One of the disks kept coming back with "electrical failure: 90% completed" on every test - though the drive still worked (between crashes).

Now, since it was new and I had a RAID backup, I am swapping it out and returning it.
But I am curious about what the nature of the problem is since googling both on this forum and elsewhere shows only rare (and unhelpful) mentions of this issue.

So any experiences or knowledge in the spirit of trying to learn something...

Thanks

[Doomer]

But I am curious about what the nature of the problem is since googling both on this forum and elsewhere shows only rare (and unhelpful) mentions of this issue.

Unfortunately the cause is vendor specific it is doubtly you will find out the real cause
"Electrical failure" it is just explanation of one of the result codes of SMART tests. This explanation was taken from ATA standard but SMART itself is mostly vendor specific thing, so that result code could be anything

[Odiferous]

I bought several new Seagate ST31000340AS drives for use in a RAID array.
After a couple of days, I kept getting sudden Linux system crashes (despite the RAID array) and independent of whether I booted from the new boot array or an old system disk.

I then ran smartctl (SMART) test on the various disks.
One of the disks kept coming back with "electrical failure: 90% completed" on every test - though the drive still worked (between crashes).

Now, since it was new and I had a RAID backup, I am swapping it out and returning it.
But I am curious about what the nature of the problem is since googling both on this forum and elsewhere shows only rare (and unhelpful) mentions of this issue.

So any experiences or knowledge in the spirit of trying to learn something...

Thanks

The problem is probably down to the fact that you are using Seagate drives.

We advise our clients not to use them, as their reliability is questionable - despite being offered with a five year warranty.

Duncan

[bnice]

I think SMART isn't SMART at all, very poorly designed and implementation, without thinking how a user would view it.

Ask a customer if they know what the SMART values are telling them, about their drive. Such a reply 'i have no idea at all, nearly all of them have high values, is that why the drive stopped working !'.

Is there resoning to why SMART hasn't been updated to correct the values the right way around, and correct its other numerous bugs. For the user to look at it, and instantly understand at once, what the drive is telling them. Not easy in educating people that SMART vales low numbers=bad and high numbers=good and the meanings of current, worst, threshold and data.