Disk & Filesystem Management (Du/Df/Lsblk/Fstrim)

Stop hunting for SSD health before your drives fail

🧩 The Challenge

You’re staring at a production server that’s throwing I/O wait spikes, but the drive isn’t reporting a full failure yet. Trying to figure out if your SSDs are actually reaching the end of their write endurance without installing a bloated monitoring agent is a nightmare.

💡 The Fix

You can pull the drive’s internal health metrics directly from the smart subsystem to see the percentage of life used. It’s way faster than guessing when you need to order a replacement.

smartctl -A /dev/sda | grep -i "percentage"

⚙️ Why It Works

The attribute 202, labeled as Percent_Lifetime_Remain or similar, gives you a raw count of how much of the NAND flash’s estimated life is still left. Seeing a low number here is usually the only warning you’ll get before things go south.

🚀 Pro-Tip: Toss this into a small loop across all your mount points found via lsblk to audit a whole rack in seconds.

Linux Tips & Tricks | © ngelinux.com | 7/17/2026