Q: During an incident, you notice that a specific log file (`/var/log/app.log`) is growing at 5GB per minute, threatening to fill the disk. Deleting the file doesn't free up the disk space space. Why, and what do you do?
In Linux, when you rm a file, you delete the directory entry (the link to the inode). However, the disk blocks are not freed as long as any running process holds an open file descriptor to that file.
Want to master this scenario in a live sandbox? The Linux Foundation's Certified System Administrator (LFCS) & SRE Track covers this exact problem with hands-on terminal drills.
🛠️ Production Runbook & Step-by-Step Resolution
Direct Answer: "During an on-call shift, our alerts triggered when a critical Linux production server exhibited this behavior. The interviewer is testing: How Linux handles open file descriptors, disk space recovery without downtime.. I structure my answer around systematic triage first, root cause analysis second, and permanent remediation third."
Initial Diagnostics & Root Cause Analysis
In Linux, when you rm a file, you delete the directory entry (the link to the inode). However, the disk blocks are not freed as long as any running process holds an open file descriptor to that file. The application is still writing to the deleted file's inode.
- Find PID using
lsof. - Find the FD number in
lsofoutput (e.g., FD 4). - Truncate it:
> /proc//fd/4
Remediation & Permanent Safeguards
Running lsof | grep deleted will show the application holding the file open. To actually free the space without restarting the application (which might cause downtime), I would truncate the file instead of deleting it. If the file is already deleted but held open, I would find it via the proc filesystem and truncate it: In the future, I'd configure logrotate to use copytruncate or send a SIGHUP to the app to force it to reopen logs gracefully.
- Find PID using lsof.
- Find the FD number in lsof output (e.g., FD 4).
- Truncate it: > /proc//fd/4
Senior hiring managers and bar raisers evaluate how you prioritize during outages. Avoid these common anti-patterns:
- Issuing immediate kill -9 (SIGKILL) commands rather than allowing graceful termination via kill -15 (SIGTERM) when configuring incident notice specific SIGKILL immediately halts execution without allowing processes to flush memory buffers, commit database transactions, or remove stale PID locks.
-
Overlooking file system inode exhaustion (df -i) when diagnosing 'No space left on device' errors when configuring incident notice specific
Millions of small files or unrotated session tokens can exhaust available file system inodes even when standard disk storage (
df -h) shows ample free space. - Exhausting open file descriptor allocations (ulimit -n) under intensive concurrent socket connections when configuring incident notice specific Reaching file descriptor exhaustion throws 'Too many open files' errors, preventing services from accepting new TCP client socket connections.
-
Permitting defunct zombie processes to accumulate without fixing parent process SIGCHLD waitpid reaping when configuring incident notice specific
Zombie processes consume process table entries; accumulating zombies can exhaust the system maximum PID limit (
pid_max).