Is knowledge in Machine Learning needed as a sysadmin??

Does a Linux Sysadmin have to know how to program/create models in Python? I understand one of the things they need to know is automating log processing. But do they have to understand how to program/create the model itself? Or are there tools already that do this?

I've done sysadmin work, the most programming you will be doing is automating tasks with powershell/python. Knowing how to program would help. But, you will almost never need to know machine learning.

>>63691 Thanks for the reply! You say I "almost" never need to know it, what are the instances where I do, if any?

>>63692 >what are the instances where I do, if any? Never (as sysadmin).

>>63692 if you work for a company that does machine learning, a high level understanding of the very basics might help you design systems that are useful for the ml team, since their requirements are occasionally different from typical software developers (their development process is more oriented around working with data sets rather than code). this is more relevant to DevOps or MLOps than sysadmin work though.

>>63690 sysadmin/it work tends to go hand in hand with a hate for programming, the most you'd do is tape together some scripts or services. if it is truly a sysadmin job and not one of the devops/doeverythingforonesalary jobs you should never run into having to do anything more than basic scripting

>>63730 Any thoughts on why that is? I'm a helpdesk veteran and I've tried to get into coding -- I wanted to be a programmer when I was 10 and I never fully let go of the fantasy of conjuring the spirits of the computer with my spells. And it's not that I can't. I can granularize problems, visualize flows, write algorithms, understand and implement data and control structures... I just can't bring myself to give any focus to learning programming in a vacuum, only when I arrive at an obstacle and can't find an existing tool to deal with it. Besides, the result is never the kind of thing a software developer does, it's all scripts and overwrought sed/awk/bash one-liners that I spent hours on just to make it concise and portable for nobody, anyway. And when I do happen upon the need to do so, I enjoy the process of picking through the documentation of a language/library and piecing together a solution for whatever dumb bullshit I decided to make a project out of that day. But it never goes deep enough or happens frequently enough to bridge the gap. I don't see how I can ever go from being an over-experienced T3 support dweeb who should be a sys/net admin by now, to being a programmer who creates comprehensive and substantial applications that solve a problem for more than two or three people at a time. Maybe it all falls into place when you bullshit your way into a job and get some tickets in front of you instead of being overwhelmed by an imaginary big picture.

>>63753 >Any thoughts on why that is? Doing the same thing over and over, as with any job. Then, adding AI development onto doing sysadmin work.

sysadmin here, lol no ML has like zero relevance to sysadmin unless maybe you're doing sysadmin for a big data company