Q: You are investigating an application crash, but no core dump was generated. How do you ensure core dumps are created for future crashes?
By default, core dumps are often disabled in production environments due to disk space and security concerns. To enable them:
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: "We encountered this OS-level bottleneck during peak traffic and diagnosed it down to kernel and filesystem metrics. The interviewer is testing: Core dump mechanisms, ulimits, systemd.. I structure my answer around systematic triage first, root cause analysis second, and permanent remediation third."
Initial Diagnostics & Root Cause Analysis
By default, core dumps are often disabled in production environments due to disk space and security concerns. To enable them:
- ulimit: Check and set the soft/hard limits for core file size.
ulimit -S -c unlimitedin the script starting the app, or edit/etc/security/limits.confto set* soft core unlimited. - Kernel Pattern: Check where dumps are written via
sysctl kernel.core_pattern. It should point to a valid directory or a handler like systemd-coredump (e.g.,|/usr/lib/systemd/systemd-coredump %P %u %g %s %t %c %h). - Systemd: If run as a systemd service, the unit file must have
LimitCORE=infinityin the[Service]section.
Remediation & Permanent Safeguards
- App configuration: Ensure the application itself doesn't catch the SEGV signal without re-raising it, or hasn't called
prctlto disable dumpability (e.g.,PR_SET_DUMPABLE).
- ulimit: Check and set the soft/hard limits for core file size. ulimit -S -c unlimited in the scri...
- Kernel Pattern: Check where dumps are written via sysctl kernel.core_pattern. It should point to ...
- Systemd: If run as a systemd service, the unit file must have LimitCORE=infinity in the [Service]...
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 investigating application crash 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 investigating application crash
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 investigating application crash 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 investigating application crash
Zombie processes consume process table entries; accumulating zombies can exhaust the system maximum PID limit (
pid_max).