Identify the system

Before changing an unfamiliar machine, identify what you are connected to. A hostname in the prompt is not enough. Two servers can have similar names and very different operating-system versions.

Start with the operating-system release:

cat /etc/os-release

Then collect the kernel, architecture, hostname, current account, and uptime:

uname -r
uname -m
hostnamectl
whoami
uptime

These commands answer different questions:

• /etc/os-release identifies the distribution and release.
• uname -r shows the running kernel.
• uname -m shows the hardware architecture, such as x86_64 or aarch64.
• hostnamectl shows the configured hostname and system details.
• whoami confirms which account will run the next command.
• uptime shows how long the machine has been running and its load averages.

Why this matters

A package name can differ between releases. A downloaded binary built for x86_64 will not run on an ARM server. A reboot may have loaded a different kernel than the one you expected.

Also verify the target before a destructive or administrative command:

printf 'host=%s user=%s\n' "$(hostname)" "$(whoami)"

You should see something like host=webserver-01 user=ada. Run that one line before rm -rf, before reboot, before any bulk package change. It takes two seconds and prevents the classic mistake of running the right command on the wrong server. Do not include secrets, private IP addresses, or customer data in a public bug report. Include only the system facts needed to reproduce the problem.

Create a small diagnostic block

When asking for help, a useful report might contain:

Ubuntu: 24.04 LTS
Kernel: 6.x
Architecture: x86_64
Service: notes-api.service
Started failing: after package update

The exact versions will vary. The important part is replacing assumptions with observed facts.

Try this on one Linux system: identify it with the commands above, then explain which command proves the distribution and which proves the architecture.