DEV Community: Muralidhar M Pala

Linux Mastery in a Line: Advanced One-Liners for Busy Sysadmins

Muralidhar M Pala — Sun, 06 Jul 2025 07:41:57 +0000

Tired of the usual df -h and uname -a? Here are 10 advanced one-liners that help real sysadmins diagnose, audit, and monitor faster.

1️⃣ Find files modified in the last 7 days, larger than 100MB:

 find / -type f -mtime -7 -size +100M 2>/dev/null

2️⃣ Check top 10 memory-consuming processes with live updates:

 watch -n 1 "ps aux --sort=-%mem | head"

3️⃣ List all listening ports with associated processes:

 lsof -i -P -n | grep LISTEN

4️⃣ Get your server’s public IP without third-party scripts:

 dig +short myip.opendns.com @resolver1.opendns.com

5️⃣ Check inode usage across filesystems (useful for mail servers):

 df -i

6️⃣ Find zombie processes:

 ps -eo pid,ppid,state,cmd | awk '$3=="Z"'

7️⃣ Quickly check CPU usage per core:

 mpstat -P ALL 1 3

8️⃣ Check which processes are using swap:

 for pid in $(ls /proc | grep -E '^[0-9]+$'); do awk '/VmSwap/{print $2 " kB\tPID=" '"$pid"'}' /proc/$pid/status 2>/dev/null; done | sort -n

9️⃣ Find largest directories in /var (top 5):

 du -Sh /var | sort -rh | head -n 5

10️⃣ Live bandwidth usage per interface:

 ifstat -t

Which of these do you use, or do you have your own hidden gems? Drop one below to share with fellow admins.
Stay tuned for next week’s Sysadmin Sunday tip to keep sharpening your command-line edge

ShellGPT for Sysadmins: Master These Lesser-Known CLI Tricks

Muralidhar M Pala — Mon, 30 Jun 2025 10:24:51 +0000

Save time and mental strain by letting ShellGPT generate clean one-liners for advanced system checks.

Being a Linux sysadmin often means juggling memory usage checks, digging through logs, finding who’s hogging ports, or debugging network issues—all while managing daily operational pressure.
Many of these CLI tasks are powerful but get avoided or deferred because remembering exact flags feels like mental tax.

Enter ShellGPT.

It generates clean, ready-to-run one-liners for advanced system checks, helping you troubleshoot and learn simultaneously.

Here’s how I use ShellGPT in my RHEL 9.6 VM lab for real-world sysadmin workflows:

1️⃣ Filter System Logs with journalctl
🔹 Prompt:

  sgpt "Write a one-liner to filter system logs for sshd failures in the last 24 hours"

Why: Saves you from recalling time flags and filtering syntax during SSH troubleshooting.

2️⃣ Find Who’s Using a Port with lsof
🔹 Prompt:

  sgpt "Generate a command to find which process is using port 443"

Why: Useful when services clash on critical ports.

3️⃣ Advanced Network Checks
🔹 Prompt:

 sgpt "How to check all active TCP connections with remote IPs and their states"

Why: Useful during connection debugging or tracing suspicious IPs.

4️⃣ Monitor Memory Usage Over Time
🔹 Prompt:

 sgpt "Create a one-liner to monitor free memory every 10 seconds with timestamp"

Why: Helps during incident analysis or load testing.

5️⃣ Disk Space Alerts
🔹 Prompt:

 sgpt "Write a bash script to alert if disk usage crosses 85%"

Why: Proactive monitoring to prevent downtime from full disks.

Why This Matters

✅ Frees your mental bandwidth to focus on architecture and higher-level tasks.
✅ Reduces repetitive Google searches for syntax.
✅ Becomes your practical LLM co-pilot for real workflows, not just theoretical exploration.

Try It Yourself

Copy-paste these prompts in your ShellGPT setup and let it simplify your sysadmin workflows.

If you find this helpful:
✅ Follow for more ShellGPT + Linux productivity workflows.
✅ Share with fellow sysadmins to save their time.
✅ Comment below on how you use ShellGPT in your lab.

✍️ Murali Pala

📍 Ajman, UAE | Ex-IAF | Linux & AI Infra | Writer

🛠️ Building AI-powered sysadmin workflows, one prompt at a time.

LinkedIn | Medium

📢 Follow for more ShellGPT and Linux productivity workflows.

ShellGPT in Cybersecurity: Your Terminal’s Silent Sentinel

Muralidhar M Pala — Mon, 23 Jun 2025 02:05:41 +0000

“As a Linux sysadmin and DevOps engineer, I juggle vulnerability scans, config hardening, and alerts—while always fighting with browser tabs.”

Cybersecurity is no longer just for pen testers or red teams. Today’s sysadmins, DevOps engineers, and solo IT warriors need to stay ahead of threats while keeping systems clean, lean, and secure. ShellGPT, a terminal‑based AI assistant, is quietly becoming a powerful ally in that mission.

In this episode, let’s explore how ShellGPT fits right into your daily security routine — helping you harden systems, audit configs, and spot trouble before it bites.

Vulnerability Scanning: nmap ...

SGPT assisting with Prompts ShellGPT can help you:

Generate custom nmap commands for quick and deep scans

sgpt " Create an nmap command to scan all TCP ports with service detection and save it to a file "

Password Policy Checks: lock it down

Weak Links Exposed

Tighten your local user security by prompting ShellGPT to:

Suggest strong password policies for /etc/login.defs and pam.d configs

Auto-create a script that flags weak passwords using cracklib-check

List audit steps for shadow file integrity

sgpt "Suggest a script using cracklib-check to audit /etc/shadow for weak user passwords and log usernames"

Firewall & Port Rules: Build with Confidence

From iptables to ufw and firewalld, you can:

Ask ShellGPT to draft precise rules (e.g., allow 443/tcp only from a subnet)

Translate abstract policy logic into shell syntax

Check for common mistakes like conflicting ACCEPT/DROP orders

sgpt  "Generate a john the ripper command to audit /etc/shadow-combined file using wordlist mode with rules"

SSH & Remote Access: Gatekeeping the Gatekeeper

One of the biggest attack vectors is open SSH. ShellGPT can:

Review your sshd_config and suggest key-based auth, limited users, port changes

Script a checker that looks for brute-force attempts in /var/log/secure

Offer fail2ban regex filters tailored to your distro

sgpt  "Give me a grep command to find recent failed SSH login attempts in /var/log/auth.log"

Log Analysis & Alerting: Parsing with Precision

Need to trace suspicious activity?

Ask ShellGPT to write grep, awk, or sed pipelines for /var/log/auth.log, /var/log/audit/audit.log, etc.

Create a script to send real-time alerts on unusual login times or IPs

Use it to build test filters for journalctl -xe

sgpt  "Suggest an awk one-liner to detect 'Failed password' patterns in /var/log/secure, outputting timestamp and IP"

Wrapping Up

ShellGPT isn’t just a shortcut — it’s a smart assistant that helps enforce good habits, spot holes, and write clear, accurate commands. Used in a controlled, ethical way, it can elevate your Linux security game without turning you into a hacker.

In our next article, we’ll take ShellGPT deeper into the world of ethical pentesting — the red team side. For now, stay safe, stay sudo.

Tell me how you're using ShellGPT—security flows, automation, or anything else you’ve built

Follow for more ShellGPT tricks and real-world command line wizardry.

10 Real-World ShellGPT Prompts Every Linux Admin Should Know

Muralidhar M Pala — Sun, 15 Jun 2025 07:41:47 +0000

If you've read my ShellGPT episodes (1, 2, and 3), you already know this isn't hype — it's hands-on. And today, for Sysadmin Sunday, I'm doing something different. No long story, no theory. Just 10 powerful ShellGPT prompts I’ve personally used as a Linux sysadmin — tested in real-world servers, not lab fantasy.

Let’s dive straight in.

1. 🔍 Check System Health

"Check CPU, RAM, disk usage, and top processes in a single-line command."

sgpt "Provide a Bash one-liner to display CPU load, memory usage, disk usage, and top 5 processes."

2. 🧠 Summarize a Log File

"Scan and summarize the last 100 lines of a syslog or journalctl output."

sgpt "Summarize the last 100 lines of /var/log/syslog to identify any warnings or errors."

3. 🔐 Audit User Access

"Find all users with sudo access and show last login details."

sgpt "A service is failing with exit code 1. Analyze the output of 'systemctl status ' and suggest a fix."

4. 📦 Troubleshoot Failing Service

Analyze systemd output and suggest possible fixes.

sgpt "A service is failing with exit code 1. Analyze the output of 'systemctl status ' and suggest a fix."

5. 🧹 Clean Up Disk Space

Suggest safe commands to clean temp files, journal logs, old kernels.

sgpt "Suggest commands to safely clean up disk space on a RHEL system, including temp files and journal logs."

6. 🗂️ Parse and Extract Logs

"Write awk/sed scripts to extract error patterns from log files."

sgpt "Write an awk command to extract all lines with ERROR from /var/log/app.log and count unique errors."

7. 🌐 Identify Open Ports and Their Services

"List all open ports and matching process names."

sgpt "Write a command to list open ports and associated process names in Linux."

8. 🕒 Monitor Real-Time Resource Usage

"Use dstat or vmstat in combination with grep to filter critical data."

sgpt "Create a real-time command using dstat to monitor CPU, disk IO, and network usage."

9. 🧱 Create and Mount LVM

"Walk through creating a volume group, logical volume, and mounting it."

sgpt "Show steps to create a new LVM from /dev/sdb1, format it, and mount to /mnt/data."

10. 🔄 Schedule Automated Backups

"Generate a cronjob to rsync user directories nightly."

sgpt "Write a cronjob that runs rsync daily at 1AM from /home to /backup."

🔗 Want More?

If you liked these prompts, check out the full episodes:

Tried one of these prompts? Let me know in the comments which worked best for your setup. And remember — the real magic is not in knowing, but in using.

Happy Sysadmin Sunday!

Follow me here on Dev.to for more ShellGPT use cases, Linux tips, and sysadmin stories from the trenches.

— Murali Pala

🔨 When Red Hat Trains Your Hands: A Walkthrough with CLAD( Command Line Assitant Daemon)

Muralidhar M Pala — Sun, 08 Jun 2025 06:18:21 +0000

After exploring ShellGPT in action, it's time to meet its quieter cousin from the enterprise world: RedHat CLAD (aka Lightspeed). Built directly into RHEL 9.6+, CLAD brings AI guidance tailored for sysadmins working in the Red Hat ecosystem. But is it just a documentation lookup engine or something more? Let's look at

🔑 What is CLAD?

CLAD (Command-Line Assistant Daemon), branded as "RHEL Lightspeed," is RedHat's built-in assistant designed to help sysadmins right inside the terminal. Unlike general-purpose AI tools, CLAD speaks RedHat. It pulls guidance aligned with best practices from official docs and IBM Watson backend.

⚙️ Real-World Admin Tasks with CLAD

Let’s walk through some common Linux sysadmin duties and see how CLAD performs:

👥 1. User Creation
Prompt: How do I create a user with a specific UID and home directory?

CLAD Response: Clear 5-step breakdown, even includes best practices and warnings.

Verdict: Like having a RedHat trainer whispering over your shoulder

📀 2. LVM Setup
Prompt: Steps to create a logical volume using LVM?

CLAD: Walks through pvcreate, vgcreate, lvcreate, and mount steps with optional flags.

Verdict: Great for brushing up. Verbose but valuable.

🚧 3. Firewall Configuration
Prompt: Allow only HTTP and SSH through firewalld?

CLAD: Sets up the public zone, adds services, checks runtime config. Also dives into internal zone usage.

Verdict: A bit more than asked, but useful if you're building secure setups.

⏰ 4. Cron Job Setup
Prompt: Run a script daily at 2 AM?

CLAD: Offers classic cron syntax with brief explanation of each field.

Verdict: Useful refresher. Adds clarity to routine tasks.

🔐 5. SELinux & Apache Troubleshooting
Prompt: How to allow blocked actions by SELinux for Apache?

CLAD: Checks audit logs, uses sealert, audit2allow, and reinforces labeling practices.

Verdict: A strong guide that keeps sysadmins out of "just disable SELinux" land.

🌟 Bonus: DevOps Quickfire

Ansible Playbook: Installs Apache, enables firewalld, copies conf file.

Terraform: CLAD walks through an EC2 setup using RHEL AMI and AWS provider.

Docker + Nginx: Steps for reverse proxy container with firewalld and networking tips.

🕵️ Final Take

CLAD is not trying to be ShellGPT. It doesn’t chase speed or versatility. It focuses on depth, safety, and alignment with Red Hat principles.

If you're a Red Hat admin or new to sysadmin tasks, it provides a great comfort layer. But if you're working across distros or need multi-tool flexibility, you might find it slow or verbose.

That’s why I keep both: ShellGPT for speed, CLAD for grounded confidence.

Want to see screenshots + step-by-step commands?

Want to see screenshots + step-by-step commands?
Check out the full walkthrough here → Medium Article

Up Next: In Episode 4, we dive into ShellGPT’s cybersecurity chops — log parsing, threat detection, and hardening made easy. Stay tuned!

Project: ShellGPT Series – AI for Sysadmins

Muralidhar M Pala — Tue, 03 Jun 2025 06:11:21 +0000

🛠️ Project: ShellGPT Series – AI for Sysadmins

I'm currently writing a multi-part series on ShellGPT, a powerful AI-driven CLI assistant that's transforming the way Linux admins and DevOps engineers work.

🔍 What it's about:

Using ShellGPT to reduce manual command-line effort
Automating routine tasks with AI
Comparing ShellGPT with Red Hat's new CLAD (Lightning Speed)
Real-world use cases in DevOps, scripting, cybersecurity, and troubleshooting
This series explores how ShellGPT — the AI-powered terminal companion — can save hours for system admins, automate repetitive tasks, and redefine DevOps workflows

📚 Episodes so far:

🔜 Coming soon:

⚔️ ShellGPT vs. CLAD (Red Hat Lightning Speed)
🛡️ ShellGPT in cybersecurity use cases
📅 Weekly micro-lessons via "System Admin Sundays" and "Terminal Tuesdays"

Follow me here or on LinkedIn for more drops. Always open to feedback and geeky convos.

Terminally in Love: Two Decades of Linux, One Shell at a Time

Muralidhar M Pala — Sun, 01 Jun 2025 17:11:25 +0000

👋 The Beginning

It all started in 2005 — RHEL 3.0 and Oracle 8.x were the giants of the day.I wasn’t just installing software — I was absorbing a way of life. I still remember setting up BIND (probably Redhat 9 (not EL)) and attending a demo at Indian Air Force HQ, Subroto Park, New Delhi. My uniform had changed. But my mindset remained tactical — observe, adapt, deploy.

Back then, there was no Stack Overflow, no YouTube how-tos. Just man pages, printed guides, and logs. And somehow, that made the learning stick deeper.

🧠 What Made Linux Stick?

Simple. The terminal felt like home.
Every dmesg, every tail -f /var/log/messages, every stubborn NIC issue — I lived through it, solved it, and learned from it.

I wasn’t chasing certs (though I earned my RHCE 4 in Dec 2005). I was chasing clarity — why a service failed, why a bootloader broke, how to recover from a corrupted /etc/fstab at 2 AM without breaking a sweat.

🧰 My Toolkit Has Changed, But Not My Spirit

Over 20 years, the tools evolved:

From init.d to systemd

From ipchains to nftables

From hand-edited xorg.conf to Wayland debates

From legacy partitions to LVM, and now ZFS experiments

And yet, I still enjoy firing up a terminal in a fresh VM more than any fancy web dashboard.

I’ve played with:

Docker, Podman, and LXC

Ansible, Terraform, and even ShellGPT recently

Monitoring stacks like Nagios → Zabbix → Prometheus + Grafana

Built my own VM lab, sometimes more powerful than bare-metal servers

🤕 A Pause, Not an End

In between, life threw a curveball. A stroke. Partial speech disability. And yet, I’m still here — typing, building, debugging, and now... sharing.

If anything, it made me more in love with Linux. The silence of the shell, the logic of grep, the clarity of top, they all became my recovery zone.

❤️ The Love Story Continues

Now I’m writing from a Ajman, UAE. A simple room. But inside it? A world of Linux boxes humming in virtual labs, prompt engineering experiments, and that same curiosity from 2005.

Two decades in — and I’m still terminally in love.

💬 Let me know if this story resonates. And if you're just starting your Linux journey, remember: you don’t need to master everything. Just don’t stop exploring the man pages.

Thanks for reading — feel free to drop a comment or connect if Linux has shaped your story too