Diagnosing Deployment Failures: Using Linux Disk Commands to Find Full Partitions

During a test deployment using a web publishing platform, one of two nodes failed to start, causing the TPS to remain low. The deployment record showed SUCCESS, but the log revealed that the failed node could not run because the disk was full.

Error log:

Command to view overall disk space:

df -h

If you only need the usage of the current directory, use:

du --max-depth=1 -h

The -h flag makes the sizes human‑readable (adds K, M, G units). The output lists the size of each subdirectory on the first n‑1 lines, with the total size on the last line.

To get the size of a single file or directory, run:

du -sh

Another equivalent command is:

du --max-depth=0 -h

In the manual, --max-depth=0 is described as having the same effect as the -s option.

You can also specify a path to check a particular location, for example:

du -sh /root/sw