Linux Command Line Essentials

09:00 – 10:15

Find your way around an unfamiliar server

the filesystem layout, which and find and locate, man pages

Takeaway A working tour of a server you have never seen: configuration, logs and application located without a browser.

10:30 – 11:45

Fix a permissions failure from the error alone

users and groups, chmod and chown, umask

Takeaway A service that cannot write its own log file diagnosed and fixed from the error message.

12:00 – 13:00

Stop a runaway process the right way

ps and top, signals, load average and memory

Takeaway A runaway process identified, stopped gracefully and investigated, with load average finally explained.

14:00 – 15:15

Answer questions from a 100MB log in one pipeline

pipes and redirection, grep and awk and sort

Takeaway The busiest endpoints by error rate pulled from a hundred-megabyte access log in a single pipeline.

15:30 – 17:00

Edit configuration on a live service, safely

vim survival, backup and validate, service reload

Takeaway A config file changed in place: backed up, validated and reloaded without downtime.

09:00 – 10:15

Diagnose a service that will not start

systemd units, journalctl, enable vs start

Takeaway A dead service diagnosed entirely from journalctl, fixed, and made to survive a reboot.

10:30 – 11:45

Work through the four classic incidents

full disk, port conflict, DNS failure

Takeaway A full disk, a port conflict, a DNS failure and a locally-fine-but-unreachable service: all four worked to the cause.

12:00 – 13:00

Set up SSH the way professionals use it

keys and agents, jump hosts, disabling password login

Takeaway Passwordless access through a jump host to two machines, with password login turned off behind you.

14:00 – 15:15

Write a shell script that survives real input

arguments and exit codes, set -euo pipefail, quoting

Takeaway A log triage script that handles errors, then fed a filename with a space in it, and surviving.

15:30 – 17:00

Rescue a server you have never seen

a broken box, a time limit, a runbook

Takeaway A genuinely broken server diagnosed end to end, and the one-page runbook you write afterwards is yours to keep.