As hard drives have become more and more commoditized and prices per TB continue to drop, manufacturers are under immense pressure to find ways to differentiate their drives and establish better profit margins. While drive consumers simply expect the drive to be fast and reliable, some manufactures seem to think that reliability is a value added option and only certain drive models should be able to run all the time without issue. Case in point is Seagate.
To confirm the data cited below and published by Backblaze, we are finding that Seagate consumer drives are extremely unreliable once they pass their 'power-on hours' rating. In what seems to be planned obsolescence, we are seeing extremely high failure rate with seagate drives when used in always on environments. I have no doubt Seagate would be the first to point out that they have more expensive drives that are intended for always-on environments, but other manufacturers' consumer drives are far more reliable as are older Seagate drives.
What's more, the failure point is so predictable that it's hard not to presume the failure point is predetermined to differentiate their consumer and enterprise drives. So it looks like the Seagate engineers forgot to randomize the failure point. Alternatively, Seagate did acquire Maxtor, who had horrendously unreliable drives at the point they were acquired, so it's quite possible that the consumer Seagate drives are just rebranded Maxtor drives.
The good news is we have had very good results with HGST and Toshiba consumer drives in our cloud servers. As of this publications we use exclusively HGST for our mechanical drive needs and our annual failure rate is once again negligible. Hopefully there is a lesson to be learned here, perhaps it's that Seagate should find some other way to differentiate their enterprise product that allowing their consumer drives to fail at such high rates.