Determine filesystem capacity utilization and headroom before hitting full disk alerts.
Filesystem utilization is simply used blocks divided by total blocks, matching the Use% column reported by `df`. Because filling a filesystem completely can cause application failures, database corruption, or an inability to even write new log entries, most monitoring conventions apply staged alert thresholds — a soft warning at 80%, a high-priority alert at 90%, and a critical/page-now alert at 95% — giving operators time to react before the filesystem is completely full.
Utilization percentage
used% = used_blocks / total_blocks × 100
Filesystems perform worse and fragment more as they approach full capacity, and many systems need free space for temp files, logs, and journaling even during normal operation — so 80% leaves a safety margin to investigate and remediate before things degrade further or writes start failing outright.
Most Linux filesystems reserve a small percentage of blocks (commonly 5% on ext4, via tune2fs -m) for the root user, so a non-root process can see 'no space left' before the raw block count hits zero.
No — a filesystem can report plenty of free blocks but still be unable to create new files if it has run out of inodes (common with filesystems storing huge numbers of tiny files); check inode usage with `df -i` separately.