Chadzilla

My go-to Linux commands when I'm troubleshooting issues in production 1. du -xh /var --max-depth=1 | sort -h | tail -5 Finds the biggest folders eating your disk. Use it when the disk is full and you need to find out who is responsible. 2. lsof +L1 | grep deleted Lists deleted files that are still held open by a process. Use it when the disk shows 100% full but du shows much less. Restart the process to get the space back. 3. ss -tulnp Shows which ports are listening and which process owns them. Use it when the service is "running" but nobody can connect to it. 4. nc -zv <host> <port> Tests if a specific port is reachable from where you are. Use it when SSH or DB connections fail. Succeeded means the network is fine. Refused means nothing is listening. Timed out means a firewall or security group is blocking. 5. dmesg -T | grep -i -E "oom|error|fail" Shows kernel messages about memory, disk, and hardware. Use it when your app keeps dying and its own logs show nothing. This is how you catch the OOM killer. 6. pidstat 2 5 Shows CPU usage per process, every 2 seconds. Use it when the server is slow and top is too jumpy to tell you who the culprit is. Add -d for disk I/O. 7. iostat -xz 1 Shows disk latency (await) and utilization (%util) per device. Use it when the server feels slow but CPU is idle. High await means the disk is the bottleneck. 8. tcpdump -i any port 443 -nn Captures live network packets on a port. Use it when nc says the port is open but requests still fail or time out. You can see whether packets arrive, get a reply, or get reset. 9. strace -p <PID> Shows every system call a process makes. Use it when an app hangs or fails silently. You will see the missing file, the permission denied, or the connection that never completes. 10. pstree -p Shows processes as a parent-child tree. Use it when you see hundreds of child processes and need to find the one parent that started them. Start from the outside (nc, tcpdump), then check the server (ss, lsof), then go deep (dmesg, strace). Which command saved you in production? Drop it in the comments 👇