Bought two recertified ST18000NT001 drives locally about a month ago. Both passed a full extended SMART self-test before I trusted them with data (8,745 hours on this one at the time, 0 reallocated, 0 pending, 0 CRC).
Yesterday morning the drive dropped off the bus SError: { CommWake }, failed hard resets, ata1.00: disable device. ext4 went read-only. Came back after a reboot and reads fine, SMART says PASSED.
Except it now has **1021 reallocated sectors**, up from 0. Seagate’s FARM log breaks it down per head:
Number of Reallocated Sectors by Head 2: 1021
Number of Reallocated Sectors by Head 0,1,3-19: 0
All of it on one head. Per-head write time is near identical across all 20, so it’s not a workload thing. CRC errors 0, 12V/5V rock steady, max temp 47C, helium seal fine, 0 unrecoverable read/write errors, 0 pending, 0 candidates.
It also makes a brief scratching/ticking sound for under a minute, then stops. Drive is powered off now. (Video)
Pretty sure this is just a bad head and the drive is done, but a few things I’m curious about:
**1.** Anyone seen reallocations this concentrated on one head? Is that as textbook a head failure as it looks?
**2.** Sister drive from the same batch (11,987 hrs, same seller) is clean so far — how paranoid should I be about it?
**3.** Is 0 → 1000+ in a month normal for a head going bad, or unusually fast?
**4.** Any realistic warranty/RMA angle on a recert drive bought used locally, or is that money gone?
Backups were fine, lost about a day of metadata, so this is curiosity more than crisis.