operatingsystemlevel

Operating system-level virtualization, also called OS-level virtualization or container virtualization, is a form of virtualization that partitions a single host operating system into multiple isolated user-space instances, or containers. Each container runs its own applications and processes with its own filesystem, network stack, and IPC namespace, but containers share a single kernel with the host and other containers.

Container environments rely on kernel features such as namespaces for isolating processes, user IDs, mount points, and networks; control groups (cgroups) for resource limits; and lightweight filesystem mechanisms (such as union or overlay filesystems) to implement images and writable layers. Because no separate kernel is started for a container, start-up times are typically fast and resource overhead is lower than hardware virtualization.

OS-level virtualization is exemplified by Linux containers (LXC), Docker's container technology, as well as FreeBSD jails and Solaris Containers, OpenVZ, and others. It is commonly used to deploy microservices, develop and test software in reproducible environments, and package applications for consistent operation across different hosts.

Advantages include low overhead, fast deployment, and efficient resource utilization. Limitations include dependence on the host kernel, which means all containers must be compatible with that kernel version; potential security trade-offs compared with full VM isolation; and limitations in running different operating systems or kernels on the same host. Some workloads requiring kernel modules or custom kernel features may be unsuitable for OS-level virtualization. The concept evolved from chroot and BSD jails, with LXC emerging in the 2000s.