How do you view and manage running processes in Linux? Explain the key commands and signals used.

Question

Accepted Answer

Process management is one of the most important skills for an SRE because you need to monitor what is running, diagnose resource-hungry processes, and gracefully stop or restart services when needed.

The ps command shows a snapshot of current processes. The most common usage is ps aux, which lists all processes with details including the user, CPU usage, memory usage, and the command that started each process. For a more focused view, ps -ef shows processes in a tree-like format with parent process IDs, which helps you understand process relationships.

The top command provides a real-time, continuously updating view of processes sorted by resource usage. It shows CPU and memory percentages, load averages, and uptime. A more modern alternative is htop, which adds color coding, mouse support, and easier navigation. You can press M in top to sort by memory usage or P to sort by CPU usage.

To find a specific process, you can combine commands: ps aux | grep nginx shows all processes related to nginx. The pgrep command is a cleaner alternative: pgrep -a nginx lists nginx process IDs and their full command lines.

Signals are how you communicate with processes. The kill command sends signals to processes by PID. The most important signals are SIGTERM (15), which asks a process to shut down gracefully and is the default signal sent by kill. SIGKILL (9) forces immediate termination and should only be used as a last resort because the process cannot clean up after itself. SIGHUP (1) often tells a process to reload its configuration without restarting.

Examples include kill 1234 to send SIGTERM to process 1234, kill -9 1234 to force kill it, and killall nginx to send SIGTERM to all nginx processes. The pkill command works similarly but matches by name pattern.

Background and foreground management is also useful. You can start a process in the background with an ampersand: long-running-task &. Use jobs to list background jobs, fg to bring one to the foreground, and bg to resume a stopped job in the background. The nohup command keeps a process running after you log out.

For SRE work, understanding process states is key: running, sleeping, stopped, and zombie. Zombie processes have finished but their parent has not collected their exit status. A few zombies are normal, but many indicate a bug in the parent process.