This crate is one of the core crates of the youki workspace, and has functions and structs that deal with the actual craetion and management of the container processes.
Remember, in the end, a container is just another process in Linux, which has control groups, namespaces, pivot_root and other mechanisms applied to it. The program executing has the impression that is is running on a complete system, but from the host system's perspective, it is just another process, and has attributes such as pid, file descriptors, etc. associated with it like any other process.
Along with the container related functions, this crate also provides Youki Config, a subset of the OCI spec config. This config contains only the essential data required for running the containers, and due to its smaller size, parsing it and passing it around is more efficient than the complete OCI spec config struct.
Other than that, this crate also provides a wrapper over basic Linux sockets, which are then used internally as well as by youki to communicate between the main youki process and the forked container process as well as the intermediate process.
This crate also provides an interface for Apparmor which is another Linux Kernel module allowing to apply security profiles on a per-program basis. More information about it can be found at https://apparmor.net/.
- rootfs, which is a ramfs like simple filesystem used by kernel during initialization
- hooks, which allow running of specified program at certain points in the container lifecycle, such as before and after creation, start etc.
- signals, which provide a wrapper to convert to and from signal numbers and text representation of signal names
- capabilities, which has functions related to set and reset specific capabilities, as well as to drop extra privileges
- tty module which daels with providing terminal interface to the container process
- pseudoterminal man page : Information about the pseudoterminal system, useful to understand console_socket parameter in create subcommand
Namespaces : namespaces provide isolation of resources such as filesystem, process ids networks etc on kernel level. This module contains structs and functions related to applying or un-applying namespaces to the calling process.
Note: clone(2) offers us the ability to enter into user and pid namespace by creating only one process. However, clone(2) can only create new pid namespace, but cannot enter into existing pid namespaces. Therefore, to enter into existing pid namespaces, we would need to fork twice. Currently, there is no getting around this limitation.
Pausing a container indicates suspending all processes in it. This can be done with signals SIGSTOP and SIGCONT, but these can be intercepted. Using cgroups to suspend and resume processes without letting tasks know.