Are there development prospects?

Question

System administration

Hello.
I have such a problem. I got a job 2.5 weeks ago in a company as an IT specialist, in fact a system administrator. The first serious job after five years of study (part-time jobs were a coder and not in my specialty)
I came to the conclusion that I am slowly leaving for a swamp - there are no tasks for about 80% of the working time. It is difficult for a young specialist to find a job in St. Petersburg, more than half of the group is still looking for it.
There was a plan to educate myself, but virtual machines are nothing compared to experience with systems.
What would you suggest?
I understood one thing - the next place of work will be a company whose main activity will be the IT industry.

Answer 1

I came to the conclusion that I am slowly going into the swamp - there are no tasks for about 80% of the working time.

This is excellent: 80% of the time you can do the most interesting tasks.

It is difficult for a young specialist to find a job in St. Petersburg

This is not true. Easier than in other places in Russia with the exception of Moscow.

There was a plan to educate myself, but virtual machines are nothing compared to experience with systems.

It depends on what kind of administrator you want to become. And IMHO virtual machines are both more interesting and much more profitable (considering the possibility of freelancing).

Answer 2

But I feel a lack of specialists...
Decide in which direction you would like to develop, find vacancies on the desired topic, look at the requirements in them, and spend 80% of the time sharpening skills to meet the requirements.
Mmm .. and it’s probably worth adding that if you have a young and really promising company, then spend time on automating and organizing processes, it’s better to do it now than later, when there will be almost no free time.

Answer 3

The system administrator can find tasks for himself.
Consider what can be automated. Imagine that you have 5-10 times more workstations in your company than there are, and implement a solution that allows you to easily manage this entire fleet.
Every thing that you come up with on your knee - try to google best practice and implement what seems rational to you.

Answer 4

The issue of orchestration and DevOPS is now quite relevant, and without "deploy a bunch of virtules and containers in OpenStack with all sorts of Puppet / Chef / Ansible / Saltstack" it's just nowhere, it's especially fun with all sorts of Zookeepers Mesos and other SOA dregs.
This is where self-education and the study of existing solutions decide, because offices usually don’t allocate money for such, and what is molded a little more than completely consists of crutches and Rubik’s cubes tied with noodles. Usually no one uses ready-made, and does not bother with long-term support, which only adds headache (and not only) pain to the whole team, not just admins.
Then it’s worth trying to write your own orchestration systems yourself (golang is best suited for this), and the repository with recipes will grow very much over time. In general, you need to study all existing deployment and orchestration systems, and then think about how and what and where.