🛡️ System Administration
📚 Documentation
Click a picture to open the full guide and its video:
Ansible
Active Directory with SSSD
Troubleshooting Active Directory
authd
Ubuntu Login with Microsoft Entra ID
Samba
Postfix
Creating SSH Keys
Passwordless SSH Login
Creating TLS Certificates
OpenSSL
Systemctl
IP Command
nmtui
Set Up the Network from a Text Menu
NIC Bonding
firewalld
Zones, Services and Rich Rules
iptables
Rules, NAT and Safe Changes
nftables
Tables, Sets and Rulesets
ufw
The Uncomplicated Firewall
SELinux
Labels, Denials and Fixes
Dig
BIND
Run Your Own DNS Server
Nmap
Tcpdump
tcpdump Filters
the BPF Language
Wireshark
Htop
Ps
Lsof
Recover a Deleted File
superfile
A Terminal File Manager
Iftop
Strace
Ltrace
Perf
Nagios Core
Hosts, Services and Alerts
Checkmk
Monitoring with Auto-Discovery
Zabbix
Install, Hosts, Triggers and Alerts
Prometheus
Metrics, PromQL and Alerting
Grafana
Dashboards, Data Sources and Alerts
Loki
Log Aggregation, LogQL and Alerts
MySQL
MySQL Replication
PostgreSQL
Receive-Only Mail in Docker
SSHFS
Remote Folders over SSH
systemd Units and Timers
Disks, LVM and Filesystems
Users, Groups and sudo
Package Management Across Distros
Ansible One-Liners
Ansible Roles and Projects
Ansible Semaphore
A Web UI and API for Playbooks
Puppet
Chef
Salt
Network Troubleshooting
Nginx Guide
journalctl
The systemd Journal
rsync
Copy, Mirror and Back Up
ss
Sockets and Listening Ports
cron
Scheduling Jobs
Runbook
Slow Server / High Load
Runbook
Disk Full
Runbook
Server Won't Boot
Runbook
apt update Fails on i386
Air-Gapped Repositories for Ubuntu and Red Hat
TLS Certificate Monitoring
Where uptime is a religion and reboot is the last resort.
Why did the sysadmin bring a ladder to the server room?
Because the uptime percentage needed a mount!
Because the uptime percentage needed a mount!
Core Principles
- Backup, Verify, Repeat — 3-2-1 rule (3 copies, 2 media, 1 offsite)
- Monitor Everything — CPU, memory, disk, network, uptime
- Automate the Boring Stuff — If you do it twice, script it
- Least Privilege — Not everyone needs sudo
⚠️ Pro Tip: Never reboot a production server without checking if someone is working on it. Use who and w first.
Coming Soon
- Linux hardening guides
- Nginx and Apache configuration
- Firewall rules (iptables/nftables)
- Log analysis with journalctl
- Performance tuning
What's a sysadmin's favorite song?
"Never Gonna Give You Up" — because the uptime never gives up!
"Never Gonna Give You Up" — because the uptime never gives up!