Linux needs FIXING 1: ps -ef
nagger 100000MRSJ Visits (872)
On my largely idle POWER8 server with 20 CPU cores (160 CPU core threads), when I run the regular command:
This results in 35 screens full of pointless process stats zooming off the top of my screen. That is 86KB dumped to my screen and 99% of it never gets read - I can't even scroll back to the start.
The problem is 1270 lines are kernel processes 970 with zero seconds of CPU time and the rest less than 10 CPU seconds in a year of server uptime!
The list includes:
There is nothing I can do about these processes, nothing I can tune and there is a set of the one process for each logical CPU = CPU Core thread. In the case of kworker, 4 processes per CPU core thread.
Note: all the processes have square brackets around them like [kworker]
Let us do some maths for the fastest Linux server on the planet: IBM POWER9 E980:
This means that all day long I use: ps -ef | grep PROCNAME
PROCNAME being the process name of what I want to see.
You get largely the same effect with: ps -ef | grep -v ]$
I also propose: ps -G <name>
To save a lot of pointless typing to remove the processes we never asked for!
What do you think?
Did you know?
The Senior Kernel Design team decided to put a stop every service running its own daemon as it was a lazy programming model.