Skip to main content
In Bash and other POSIX-compatible shells, special variables like $$ and $! help you manage process IDs (PIDs) for debugging, automation, and scripting tasks. While they might not pop up every day, knowing how to use them can streamline background jobs, service management, and process monitoring. Before diving into these variables, let’s clarify how TTYs, shells, and PIDs relate:
  1. When you open a terminal, it’s assigned a TTY (teletype) name.
  2. A shell (e.g., Bash) runs on that TTY and acts as the parent process.
  3. Any command or script executed in that shell becomes a child process.
  4. Every process—from parent shells to background jobs—has a PID and a lifecycle (start → run → exit).

​
Table of Special PID Variables

VariableDescriptionTypical Use Case
$!PID of the most recently launched background jobTracking & controlling background tasks
$$PID of the current shell or shell-script processSelf-awareness in scripts

​
Using $! to Capture Background Job PIDs

The $! variable returns the PID of the last job you sent to the background. 
$ sleep 5 &
[1] 93506
$ echo $!
93506
Even if you run foreground commands afterward, $! holds the PID until you start another background job: 
$ echo "First background PID is $!"
First background PID is 93506

$ echo "Hello" &
[1] 93761
Hello
[1] + done echo "Hello"

$ echo $!
93761

​
Storing $! in a Bash Script

A common pattern is launching a service in the background, capturing its PID, and later terminating it. For example, starting an Apache JMeter server: 
#!/usr/bin/env bash

jmeter_pid=""

start_jmeter() {
  echo "Starting JMeter server..."
  jmeter-server &
  jmeter_pid=$!
}

start_jmeter

# Other workflow steps...
echo "Now running additional setup tasks..."

<Callout icon="triangle-alert" color="#FF6B6B">
Always ensure you `kill` the correct PID. Accidentally terminating the wrong process can disrupt critical services.
</Callout>

echo "Stopping JMeter server (PID: $jmeter_pid)"
kill -SIGTERM "$jmeter_pid"

​
Using $$ to Identify Your Shell or Script

The $$ variable prints the PID of the current shell or script process: 
$ echo $$
79315
$ ps -p $$ -o pid,cmd
  PID CMD
79315 -bash
Opening a new terminal tab or window yields a different $$ value: 
$ echo $$
91933

​
Inspecting $$ Inside a Script

Create print_pid.sh: 
#!/usr/bin/env bash
echo "This script’s PID is $$"
sleep 60
Run it in the background: 
$ ./print_pid.sh &
[1] 94479
94479
$ ps --pid 94479,94481 -o pid,ppid,cmd
  PID  PPID CMD
94479 79315 /bin/bash ./print_pid.sh
94481 94479 sleep 60
  • PID 94479 corresponds to the script itself ($$).
  • PID 94481 is the child sleep process.

​
$$ and Subshell Behavior

Subshells inherit the parent shell’s PID, so $$ stays constant: 
#!/usr/bin/env bash

echo "Parent shell PID: $$"
(
  echo "Inside subshell PID: $$"
)
Output: 
$ ./subshell_pid.sh
Parent shell PID: 54344
Inside subshell PID: 54344
All references to $$ show the same PID because subshells share the parent’s process ID.

​
Key Takeaways

  • $! returns the PID of the last command run in the background.
  • $$ returns the PID of your current shell or the running script.
Understanding these variables will help you write more robust scripts, automate process control, and debug complex workflows with confidence.

Watch Video