Hard drive randomly quits responding

[Planefreak]

Hello,

I've been having a persistent problem that one of my non-critical spinning drives (I have 10 drives installed, 2 SSD and 8 spinning discs) will randomly stop responding and Debian 12 will report an input/output error.

I tried using the built in disk management to unmount and mount, but I am still unable to communicate with the drive. I tried this again from the terminal using umount, but I could not get it to work. Then I removed the disk from the machine and scanned it on a different computer system and did not find any errors. No bad sectors found and I don't recall seeing anything in the SMART data that might indicate an issue.

The only thing I've found to get the drive to reconnect is to reboot my machine. This is a hassle because I have to shut down several applications and relaunch some of them after booting. The drive is always connected and working after it boots. Sometimes it will be a few hours before the drive reports input/output error and sometimes its weeks.

When this started, I swapped out the cable with a new one. Other than that, I haven't been in the server for a few months.

Where do I start with fixing this issue? Thank you


OS: Debian 12.
HDD: SATA 14TB Toshiba with around 15,000 hours. Formatted as NTFS. All data is backed up.

[mrmazda]

For what definition of "new"? Freshly purchased? Bought years ago but still in original wrapper, but may not support SATA3.x/600MB/s? I have lots of the latter. :p My luck with Toshiba spinners hasn't been good either.

[Planefreak]

I tried swapping cables with another drive, as in it was connected to a different motherboard port on a different cable. The problem persisted, so I used one the spare cables I bought when I built the server 7 months ago. I'll replace the cable now.

Been a while since I've had a drive failure. Might be the case here. Hope not.

[Stones]

Hello,

Have you considered checking the system logs for additional clues? Sometimes they can reveal valuable information. By the way, have you shared your troubleshooting experiences on Twitter? It's often a great way to get additional tips and tricks from the community."