1
1
I have bought 4 used 6TB WD Red drives (WDC WD60EFRX-68MYMN1) for use in my HP Microserver Gen8. The server is running Debian Linux 9. The disk controller is using AHCI, no hardware RAID.
Initially I tested all drives, checked SMART values using smartctl (lifetime was around 2.5 years for all of them), ran some self tests (conveyance and long), read the entire surface, and I'm sure they're healthy.
During data migration, I unplugged and plugged back in all drives several times. The Microserver advises against that with "non hot-swap" written on the drive bays, but I ignore this as it tends to work regardless.
Once the migration was done and the dust had settled, I check the SMART values again. To my surprise, 2 of the 4 drives had their SMART values literally reset. All values were back to 0, and they're happily counting back up again (e.g. in the case of power on hours or start/stop count). Other than that, they seem to be working perfectly fine.
How is this possible? If it had just been one drive, I would chalk it up to a random fluke. But two drives?
Searching the 'tubes reveals no such occurences having happened before, and while certain Seagate drives apparently had, for a time, the ability to manually reset SMART, it always involved hooking up a serial interface to the extra pins next to the SATA connector and deliberately initiating the process.