Linux

What is Linux and what does it do?

Linux is an operating-system kernel. It sits between hardware and user-space programs, schedules processes, manages memory, exposes filesystems and coordinates networking and devices. Applications request these services through system calls rather than controlling hardware directly.

The word Linux is also used informally for complete operating systems, but the kernel and a distribution are different layers. A distribution combines the Linux kernel with libraries, package repositories, service-management conventions and administrative tools.

The kernel boundary

The kernel decides which process runs, which memory pages it may access and how input or output reaches a device. File descriptors give programs a common handle for files, sockets and other resources. Permissions and capabilities determine whether an operation is allowed.

This boundary makes isolation and accounting possible. Namespaces can provide different views of processes, mounts or networks. Control groups organize processes hierarchically and distribute or measure resources such as CPU and memory. Containers build on these kernel mechanisms rather than introducing a separate guest kernel.

Processes, resources and evidence

A running service has several connected forms of evidence:

• Process identity and parent-child relationships show what is executing.
• Open files and sockets show which resources it currently uses.
• Mounts and permissions explain which data it can reach.
• Resource accounting shows pressure, limits and consumption.
• Kernel and service logs provide events that are not visible from one process snapshot.

Diagnosis is strongest when these views answer a specific question. Restarting a process may remove the immediate symptom while also discarding the state needed to explain it.

What a distribution adds

Package managers track installed files and dependencies. Service managers define startup, shutdown and supervision behavior. Distribution documentation selects supported configuration paths from the much broader set of kernel capabilities.

A kernel feature should not be confused with a distribution default. The available kernel interface, enabled configuration, package version and local policy all affect what a host can actually do.

Control comes with responsibility

Linux fits servers, development systems, embedded devices and infrastructure where inspectability and automation matter. The same composable interfaces that make small systems understandable can also support large operational environments.

The boundary is complexity. Permissions, capabilities, namespaces, control groups, filesystems and networking interact. A command copied without context can change persistent state or interrupt unrelated services. Safe operation begins with read-only evidence, continues through distribution-supported configuration and ends with small reviewed changes that have a tested recovery path.