Linux Process Priority

Linux Cgroups provide a framework to apply resource restrictions to Linux systems
Cgroups can limit the amount of CPU cycles, available memory, and more
If processes are equal from a perspective of Cgroups, the Linux nice and renice commands can be used to manage priority

In Cgroups, the Linux system is divided in 3 slices
- System: all systemd processes
- User: all user processes
- Machine: Virtual machines and containers
Each slice has an equal CPU weight
That means that if one or more processes within a slice requests a maximum amount of CPU cycles, each slice will get an equal amount of CPU shares
- So 20 systemd processes together gets as much as one user process that claims full CPU usage!
In systemd, the CPUWeight can be set on individual systemd units

nice and renice can’t change the priority if,
- specific cgroups are defined
- if the process is a realtime process (ex: Kernal Processes, these are in a different queue)

Users can set their processes to a lower priority, to increase priorities you need root access
Priority is always relative to other processes

nice -n 19 dd if=/dev/zero of=/dev/null

dd if=/dev/zero of=/dev/null:

dd is a command-line utility for copying and converting data.
if=/dev/zero specifies the input file as /dev/zero, which generates an endless stream of null bytes (zeros).
of=/dev/null specifies the output file as /dev/null, a special file that discards all data written to it.
Effectively, this setup copies an infinite stream of zeros from /dev/zero and discards it into /dev/null.