Updated: 12 October 2026 · Applies to: procps-ng 4.0.4 (ps, top, pgrep), htop 3.4.1, psmisc 23.7 and systemd 257 on Debian 13, tested; Ubuntu 24.04 and 26.04 ship the same procps-ng 4.0.4; ps, pgrep and htop 3.3.0 (EPEL) on AlmaLinux 10
To list all running processes in Linux, run ps aux. It shows every process with its user, process ID (PID), CPU and memory use and the command. For a live view that keeps the busiest processes at the top, run top (or htop). To find a process by name, run pgrep -a name.
ps aux # every process: user, PID, CPU %, memory %, command ps -ef # every process with its parent process ID (PPID) top # live view, press q to quit pgrep -a nginx # find processes by name systemctl list-units --type=service --state=running # running services
The output below is a lab example from a Debian 13 test container, so it shows only a few processes. A real server usually lists many more, including kernel threads, whose names are shown in square brackets. ps -e --no-headers | wc -l counts them.
1. List all processes with ps
$ ps aux USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND root 1 0.1 0.3 21676 12672 ? Ss 03:25 0:00 /sbin/init root 342 0.1 0.2 34444 8764 ? Ss 03:26 0:00 /usr/lib/systemd/systemd-journald root 363 0.0 0.0 4288 2432 ? Ss 03:26 0:00 /usr/sbin/cron -f root 380 0.4 0.2 14460 8448 ? Ss 03:27 0:00 /usr/bin/python3 -c import time; time.sleep(3600) alice 383 0.0 0.0 2580 1408 ? S 03:27 0:00 sleep 1000 root 392 0.0 0.0 6392 3584 ? R 03:27 0:00 ps aux
RSS is the physical memory (RAM) the process uses, in KiB. START is when it started and TIME the CPU time it has used. ps -ef lists the same processes in the other common format, with the parent PID, which tells you which process started which.
Running or sleeping? The STAT column shows the state. According to the ps manual, R is running or runnable, S is sleeping (waiting for an event), D is uninterruptible sleep (usually disk I/O), T is stopped and Z is a zombie. On a healthy server almost every process is S: it exists and works, but is waiting for something. To list only the processes that are on the CPU right now, add r. In our test that was just ps itself:
$ ps aux r USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND root 395 0.0 0.0 6392 3712 ? R 03:27 0:00 ps aux r
Sort and filter the list:
ps aux --sort=-%mem | head # most memory first ps aux --sort=-%cpu | head # most CPU first ps -u alice -o pid,etime,cmd # only the processes of user alice
Pick your own columns with -o. ELAPSED (etime) is how long each process has been running:
$ ps -eo pid,user,%cpu,%mem,etime,cmd --sort=-%cpu | head -4
PID USER %CPU %MEM ELAPSED CMD
380 root 0.4 0.2 00:02 /usr/bin/python3 -c import time; time.sleep(3600)
342 root 0.1 0.2 00:37 /usr/lib/systemd/systemd-journald
1 root 0.1 0.3 01:33 /sbin/init
More ways to find the heaviest processes: Find Top Running Processes by Highest Memory and CPU Usage in Linux.
2. Watch processes live with top
top refreshes every few seconds. Its header shows the totals, for example Tasks: 8 total, 1 running, 7 sleeping, 0 stopped, 0 zombie. Keys you will use most (from the top manual):
P: sort by CPU,M: sort by memory,T: sort by CPU time.u: show one user's processes;c: show the full command line.1: show each CPU separately.k: kill a process (top asks for the PID and the signal);q: quit.
For a one-time snapshot you can save or paste into a support ticket, use batch mode:
$ top -b -n 1 | head -14
top - 03:27:37 up 12 days, 20:58, 0 users, load average: 1.37, 0.94, 1.00
Tasks: 8 total, 1 running, 7 sleeping, 0 stopped, 0 zombie
(CPU and memory lines and the last two processes left out)
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1 root 20 0 21676 12672 10112 S 0.0 0.3 0:00.10 systemd
342 root 20 0 34444 8764 7740 S 0.0 0.2 0:00.04 systemd-journal
363 root 20 0 4288 2432 2304 S 0.0 0.1 0:00.00 cron
380 root 20 0 14460 8448 5760 S 0.0 0.2 0:00.01 python3
383 alice 20 0 2580 1408 1408 S 0.0 0.0 0:00.00 sleep
top -u alice shows one user and top -p 380 one process. More batch-mode examples: How to find top 15 Processes by Memory Usage with top in Batch Mode?
3. An easier live view: htop
sudo apt install htop # Debian, Ubuntu sudo dnf install epel-release && sudo dnf install htop # AlmaLinux, Rocky Linux (htop is in EPEL)
htop shows CPU and memory bars and lets you scroll through the list with the arrow keys. According to the htop manual: F3 or / searches, F4 or \ filters the list, F5 or t shows a tree, F6 picks the sort column, u shows one user, F9 or k sends a signal (kill) and F10 or q quits. M and P sort by memory and CPU, as in top.
4. Find a process by name
$ pgrep python3 # PIDs only 380 $ pgrep -a python3 # PIDs with the full command 380 /usr/bin/python3 -c import time; time.sleep(3600) $ pgrep -a -u alice # everything user alice runs 383 sleep 1000 $ pgrep -c sleep # how many match 1 $ pidof cron # PIDs of an exact program name 363
When nothing matches, pgrep prints nothing and returns exit code 1, which is handy in scripts. If you use grep, put the first letter in brackets so the grep command does not find itself: ps aux | grep '[p]ython3' showed only the Python process, while ps aux | grep python3 also listed grep python3.
Related guides: get a process name from a PID, or a PID from a name, and How to Find Which Process Is Using a Port in Linux (ss, lsof, fuser).
5. Show which process started which
$ pstree -p
systemd(1)-+-cron(363)
|-python3(380)
|-sleep(383)
`-systemd-journal(342)
pstree is in the psmisc package. Without it, ps -ef --forest draws the same tree.
6. List running services
Most server software (web server, database, mail) runs as a systemd service. List the ones that are running:
$ systemctl list-units --type=service --state=running UNIT LOAD ACTIVE SUB DESCRIPTION cron.service loaded active running Regular background program processing daemon report-worker.service loaded active running Report worker systemd-journald.service loaded active running Journal Service (legend shortened) 3 loaded units listed.
systemctl status report-worker shows a service's state and its Main PID (380 here), and systemctl show -P MainPID report-worker prints just the PID.
7. Stop a process
kill 380 # ask process 380 to stop (SIGTERM) kill -9 380 # force it (SIGKILL), only if it does not stop pkill -f "python3 -c" # by name or part of the command line
For a service, use systemctl stop or systemctl restart instead of kill, so systemd knows the service was stopped on purpose (a service set to restart automatically would otherwise come back). Details: How to Find and Kill Running Processes in Linux?
More ps options: ps command switches and usages and 30 useful ps command examples. For CPU and memory totals, see How to Check CPU Usage on Linux: top, htop, mpstat and sar and How to Check Memory Usage in Linux (free, top, meminfo). If your Ucartz server is busy and you cannot tell why, attach the output of top -b -n 1 to a support ticket.
