Skip to content

Job Control

Job control lets an interactive shell run commands in the background, stop them and move them between foreground and background. It also decides what happens to those commands when the terminal or SSH session closes, which is where most "my job died when I logged out" problems start.

Track: Core · Interview weight: Med

Must-Know Facts

Fact Value Verify with
cmd & Runs in the background; the shell prints [job] PID sleep 60 &
jobs -l Lists the shell's jobs with PIDs; + is current, - is previous jobs -l
Ctrl-Z Sends SIGTSTP; the job stops jobs shows Stopped
bg %n / fg %n Continue a job in the background / bring it to the foreground bg %1
Ctrl-C Sends SIGINT to the foreground process group; exit status 130 echo $?
Ctrl-D End of input (EOF), not a signal; closes the shell at an empty prompt cat then Ctrl-D
Job specs %1, %+, %-, %sleep (name prefix) kill %1
Terminal hangup The session leader gets SIGHUP; bash sends SIGHUP to all its jobs close the terminal
nohup Ignores SIGHUP; writes output to nohup.out if stdout is a terminal nohup cmd &
disown Removes a job from the shell's table, so bash does not send it SIGHUP disown %1
setsid Starts a command in a new session with no terminal setsid cmd
huponexit If set, bash sends SIGHUP to jobs on a normal exit too (off by default) shopt huponexit
tmux / screen Keep a whole terminal session alive on the server across disconnects tmux attach

Background, Stop and Resume

The session below ran in an interactive bash on a pseudo-terminal; ^Z and ^C are the keys typed.

sleep 300 &
sleep 400 &
jobs -l
sleep 500
# Ctrl-Z
jobs
bg %3
jobs
kill %1
jobs
fg %2
# Ctrl-C
echo $?
jobs
disown %3
jobs; ps -o pid,ppid,stat,comm -C sleep

Output:

$ sleep 300 &
[1] 2597
$ sleep 400 &
[2] 2598
$ jobs -l
[1]-  2597 Running                 sleep 300 &
[2]+  2598 Running                 sleep 400 &
$ sleep 500
^Z
[3]+  Stopped                 sleep 500
$ jobs
[1]   Running                 sleep 300 &
[2]-  Running                 sleep 400 &
[3]+  Stopped                 sleep 500
$ bg %3
[3]+ sleep 500 &
$ jobs
[1]   Running                 sleep 300 &
[2]-  Running                 sleep 400 &
[3]+  Running                 sleep 500 &
$ kill %1
$ jobs
[1]   Terminated              sleep 300
[2]-  Running                 sleep 400 &
[3]+  Running                 sleep 500 &
$ fg %2
sleep 400
^C
$ echo $?
130
$ jobs
[3]+  Running                 sleep 500 &
$ disown %3
$ jobs; ps -o pid,ppid,stat,comm -C sleep
    PID    PPID STAT COMMAND
   2599    2596 S    sleep

After disown, jobs no longer lists sleep 500, but the process keeps running as a child of the shell.

Key or command Effect
Ctrl-Z Stop the foreground job (SIGTSTP)
bg Resume the current job in the background (SIGCONT)
fg Resume it in the foreground
Ctrl-C Interrupt the foreground job (SIGINT)
Ctrl-\ Quit with a core dump (SIGQUIT)
kill %n SIGTERM to every process of job n
wait Block until all background jobs finish

A background job that reads from the terminal receives SIGTTIN and stops at once:

cat &
jobs

Output:

[1] 6329
[1]+  Stopped                 cat

Give background jobs their own input

A job waiting on SIGTTIN does nothing until it is brought to the foreground. Redirect stdin from a file or /dev/null for anything started with &.

What Survives a Closed Terminal

Four background jobs were started in an interactive shell, then the terminal was closed, which sends SIGHUP to the shell.

sleep 601 &
nohup sleep 602 > /dev/null 2>&1 &
setsid sleep 603 &
sleep 604 & disown
ps -o pid,ppid,sid,tty,comm -C sleep
echo '--- after the terminal closed ---'
ps -o pid,ppid,sid,tty,args -C sleep

Output:

$ sleep 601 &
[1] 2645
$ nohup sleep 602 > /dev/null 2>&1 &
[2] 2646
$ setsid sleep 603 &
[3] 2647
$ sleep 604 & disown
[4] 2649
[3]   Done                    setsid sleep 603
$ ps -o pid,ppid,sid,tty,comm -C sleep
    PID    PPID     SID TT       COMMAND
   2645    2643    2643 pts/0    sleep
   2646    2643    2643 pts/0    sleep
   2648       1    2648 ?        sleep
   2649    2643    2643 pts/0    sleep
--- after the terminal closed ---
    PID    PPID     SID TT       COMMAND
   2646       1    2643 ?        sleep 602
   2648       1    2648 ?        sleep 603
   2649       1    2643 ?        sleep 604
Job Survived Why
sleep 601 & No bash forwarded SIGHUP, and sleep has the default action
nohup sleep 602 & Yes SIGHUP is ignored
setsid sleep 603 & Yes Different session; no terminal to hang up
sleep 604 & disown Yes Not in bash's job table, so bash did not signal it

setsid forked because the shell had made it a process group leader, so the job table reported Done for PID 2647 while the real sleep ran as PID 2648 in its own session.

nohup does not protect against a closed stdout

A nohup job that still writes to the terminal gets EIO or SIGPIPE after the disconnect. Redirect output to a file, as nohup cmd > job.log 2>&1 & does.

tmux

tmux keeps shells running inside a server process that is independent of the SSH connection. It is the practical choice for long interactive work on a remote host.

Command Effect
tmux new -s deploy New session named deploy
Ctrl-b d Detach
tmux ls List sessions
tmux attach -t deploy Reattach
Ctrl-b c / Ctrl-b n New window / next window
Ctrl-b % / Ctrl-b " Split vertically / horizontally
tmux kill-session -t deploy End the session

Use a systemd unit for anything that must outlive a reboot

nohup, disown and tmux protect against a disconnect, not against a reboot or a crash. systemd-run --unit=<name> <cmd> gives a job logging, restart policy and a status command; see Systemd Toolbox.

Common Errors

bash: fg: current: no such job

Cause: the job table of this shell is empty: the job finished, was disowned, or belongs to another shell.

Fix: find the process with pgrep -a <name>; a process from another shell cannot be brought to this terminal, but reptyr <pid> can move it on some systems.

Interview Checkpoints

L1: What does Ctrl-Z do, and how do you continue the job?

Say first: it sends SIGTSTP and stops the foreground job; bg continues it in the background and fg in the foreground.

Proof: jobs shows Stopped, then Running after bg %1.

Follow-up: Which signal do bg and fg send? (SIGCONT.)

L1: Why does a background job die when the SSH session closes?

Say first: the hangup sends SIGHUP to the shell, which forwards it to its jobs, and the default action of SIGHUP is to terminate.

Proof: after closing the terminal, only the nohup, setsid and disown jobs remain in ps.

Follow-up: Which of those still dies if it writes to the closed terminal?

L1: What is the difference between Ctrl-C and Ctrl-D?

Say first: Ctrl-C sends SIGINT to the foreground process group; Ctrl-D sends end-of-file to a program reading the terminal.

Proof: sleep 100 then Ctrl-C gives $? 130; cat then Ctrl-D exits 0.

Follow-up: Why does Ctrl-D log out an empty shell?

L2: A long command is running in the foreground and the session must be left. Keep it running.

Say first: stop it, resume it in the background, and remove it from the job table.

Proof: Ctrl-Z, then bg, then disown -h %1 (or disown %1).

Follow-up: What happens to its output?

L2: Start a job that survives logout and keeps a log.

Say first: ignore SIGHUP and redirect every stream.

Proof: nohup ./backup.sh > backup.log 2>&1 < /dev/null &

Follow-up: How would you do it with systemd instead? (systemd-run --unit=backup ./backup.sh.)

L2: Kill every job the current shell started.

Say first: pass the job PIDs to kill.

Proof: kill $(jobs -p)

Follow-up: Why does this not touch disowned processes?

L3: A developer says their nohup job still stops after they disconnect.

Say first: check what the process receives and where its output goes.

Proof: grep SigIgn /proc/<pid>/status for bit 0; ls -l /proc/<pid>/fd/1 for a terminal device; journalctl --since for SIGPIPE or an OOM kill.

Follow-up: How does tmux avoid the problem?

Captured on Rocky Linux 10.2 (iximiuz Labs microVM, kernel 6.1.167), 2026-09.