ulimit

Ultilimit? The term is actually “ulimit.” It is a shell builtin in many Unix-like systems that controls the resources available to the shell and to the processes started by it. The limits exposed by ulimit reflect the kernel’s resource limits, commonly implemented via getrlimit and setrlimit, and therefore affect only the current session and its children unless changed again in a broader context.

Soft and hard limits: Each resource limit has a soft value, which the kernel enforces for the running process. A hard limit acts as an upper bound that typically requires elevated privileges to increase. Users or administrators can tighten limits to prevent runaway processes or protect system stability.

Common limits: Typical resources include CPU time (-t), file size (-f), data segment size (-d), stack size (-s), core file size (-c), number of open files (-n), resident set size (-m), virtual memory (-v), and the number of processes (-u). The exact options can vary by shell, but Bash, for example, supports a wide set of these.

Usage: The command ulimit -a shows all current limits. To set a limit, specify the option and value, possibly with -S for soft or -H for hard. Examples: ulimit -n 1024 sets the soft limit on open file descriptors to 1024; ulimit -S -n 2048 raises only the soft limit (if permitted); ulimit -t 60 sets CPU time to 60 seconds.

Scope and configuration: ulimit affects the current shell session and its children. For persistent system-wide limits, administrators use configuration files such as /etc/security/limits.conf (on Linux) or PAM modules, rather than per-shell ulimit changes.