Last Blog Review →

In the last blog we understood, the process priority can scheduled manually during it’s launch process via. nice cmd. But if a running process’s priority has to be re-scheduled then it can done via. renice command.

Today’s Agenda → Understanding what is Trap in Linux

What is trap ??

Let’s understand with a simple example. You as a Software engineer write a script, in which you create a temporary file to store the data and at the end of the script you delete the temporary file. Now when the script executes, suddenly the script receives a signal of exit say with crtl+c this leads to the un-expected end of the script execution that too in between. Now as the script is not executed completely the temporary file is present as it is in the file directory.

This will go un-notice, as the users wont be aware of the backend program and what it does. Which leads to the un-necessary file clot in the directory and consumption of storage space.

Solution ??

Trap comes in the picture. What it says is that, you as a software engineer tell me what all signal you want me (trap) to handle when your script runs and when i receive that/those signals, tell me what action i need to perform…

Syntax

trap action signals

Real-World DevOps Example: Graceful Shutdown of a Long-Running Script

🧩 Scenario

A DevOps engineer runs a script that:

• processes logs continuously

• uploads results to cloud storage

• uses a lock file to prevent multiple instances

• writes output to a temporary file

During deployment or server shutdown, the script gets a SIGTERM signal from systemd.

Without handling it, the script might:

• leave lock files behind

• leave temp files

• corrupt partial data

• break the next run

So DevOps engineers use trap to ensure a graceful shutdown.

🔧 Example: Log Processor With Graceful Shutdown

#!/bin/bash

LOCKFILE="/var/run/log_processor.lock"
TMPFILE="/tmp/log_buffer.txt"
LOGFILE="/var/log/log_processor.log"

# Create lock file
echo $$ > "$LOCKFILE"

# Handle shutdown signals: SIGINT (Ctrl+C), SIGTERM (systemd stop), EXIT (any exit)
trap "cleanup" SIGINT SIGTERM EXIT

cleanup() {
    echo "$(date): Shutting down..." >> $LOGFILE

    # Upload partial data before quitting
    if [[ -s $TMPFILE ]]; then
        echo "$(date): Uploading buffered data..." >> $LOGFILE
        aws s3 cp $TMPFILE s3://mybucket/logs/partial_$(date +%s).txt
    fi

    # Remove temporary files and lock file
    rm -f "$TMPFILE" "$LOCKFILE"

    echo "$(date): Cleanup complete. Exiting." >> $LOGFILE
}

echo "Processing logs... (Press Ctrl+C to stop). Check the log file at /var/log/log_processor.log"

# Main worker loop
while true
do
    journalctl -f | head -n 10 >> "$TMPFILE"
    sleep 1
done

🧠 What This Script Does (In Simple Terms)

✔ Creates a lock file
✔ Starts processing logs
✔ Buffers data into a tmp file
✔ If interrupted by:

• Ctrl+C

• systemctl stop

• server shutdown

• script error

• any exit

➡ It runs the cleanup function first.

Inside cleanup it:

• uploads partial data to S3

• removes lock and temp files

• logs the shutdown event

• exits safely

Conclusion →

In this blog we understood, what is a trap command. Why we use the trap command. The syntax of the trap command. And the real-world example of the trap command.