Two dead Seagate drives, one JBOD NAS, and a hard lesson in redundancy.

In one week we had a Seagate ST2000DM001, an ST2000LX001 (a hybrid SSD/HDD unit), a 1TB laptop SSHD and an ST1000LM035 across the bench — clicking Seagates, no-power Seagates, and drives that simply would not spin.

In our experience, Seagate drives are among the least forgiving mechanisms to recover from. The 1TB and 2TB backup and expansion drives fail early and often, sometimes within months of purchase. A minor knock can be catastrophic, and Seagate encrypts the firmware on many models, which adds real difficulty to a recovery that would otherwise be straightforward.

The ST2000DM001 — Seagate's 2TB Barracuda produced from 2013 onwards — has a particularly poor reputation. Published failure-rate data has put it above 35%, and when these drives fail, they tend to fail hard: partial recoveries are common, full ones are not.

Clicking is the clearest warning sign, and the pattern tells us something before we even open the drive. Three or four clicks followed by a spin-down usually points to media damage. Twelve clicks and a spin-down is much the same. Eleven clicks, particularly on the older 7200.11 series, is more often a firmware fault. A drive that keeps clicking but keeps spinning up again usually has bad blocks on the magnetic surface, or firmware that has begun corrupting itself.

Seagate also manufactures drives sold under other names — LaCie, Maxtor, and Samsung's spinning drives after 2013 are built on Seagate mechanisms internally. So "not a Seagate" on the case doesn't always mean not a Seagate inside.

The client went away for the weekend. When they came back, the NAS had stopped responding. Inside were two 2TB Seagate ST2000DM001 drives — the same model behind that 35%+ failure statistic — and both were completely dead: no spin, no noise, nothing.

The two drives were set up as JBOD, not RAID 1. That distinction mattered enormously to the outcome. In JBOD, the drives are simply pooled together with no mirroring, so each disk held data the other didn't. One drive going down wasn't an inconvenience to be absorbed by the other — it was a straight loss of whatever that specific drive held.

Opening the drives up, both PCBs had blown fuses — a common Seagate failure point, and the reason for the total silence on power-up.

Drive one — the disk holding the current business data — could not be recovered. The ring scratches on the platter proved unrecoverable even with a donor head swap, which is as far as this kind of physical damage can be pushed.

Drive two was successfully imaged and yielded 140GB of data, dating back to 2016. Not the current files the client needed most, but not nothing either.

The result underlines why JBOD is a risky choice for a two-bay NAS holding anything you can't afford to lose. With RAID 1, either drive failing would have left a full mirror on the other. With JBOD, each drive was a single point of failure for its own data — and Seagate's ST2000DM001, statistically, was never a drive to take that chance with.