Parallel processing in bash is easy with the wait command. There is a small gotcha to watch out for. To use it wait will wait until all child background processes are finished. In your bash script you can toss a bunch of work into background processes with & and instead of exiting, calling waitwill hang until they all complete:

#!/bin/bash

do_something() {
  sleep 1
  echo $1
}

for i in {1..5}; do
  do_something $i &
done

wait

echo "All done!"

Running this will output something like this after 1 second, instead of the 5 seconds needed by each function.

5
1
3
2
4
All done!

The gotcha

Now the gotcha in that example has to do with handling error codes of the child processes. The difference is wait vs wait $pid.Waiting without a pid will swallow the exit code of the child process. Wait with a pid will return the exit code, so if we want to handle errors we will need to keep track of the pids of our child processes

The bash variable$!holds the pid of the most recently created child process. Modifying our loop we can track the pids of each created process to an array:

for i in {1..5}; do
    do_something "$i" &
    pids+=($!)
done

And instead of waiting for all pids with wait we wait for each pid individually, and track their exit codes in an array:

for pid in "${pids[@]}"; do
    wait "${pid}"
    status+=($?)
done

Putting this all together we now have parallel processing while handling exit codes of child processes:

#!/bin/bash

do_something() {
    sleep 1
    echo "$1"
    exit $((i % 2))
}

for i in {1..5}; do
    do_something "$i" &
    pids+=($!)
done

for pid in "${pids[@]}"; do
    wait "${pid}"
    status+=($?)
done

for i in "${!status[@]}"; do
    echo "job $i exited with ${status[$i]}"
done

echo "All done"

The output should look something like this, with the option of doing what you need on a per exit code basis:

1
5
4
3
2
job 0 exited with 1
job 1 exited with 0
job 2 exited with 1
job 3 exited with 0
job 4 exited with 1
All done