How to understand the architecture of servers without documentation for it?

Question

linux

Good afternoon! I'm still quite new to administration, I need advice.

Such a situation:
They give me 2 servers on centos for support without any formalized documentation on them.
I was given time to figure out what is installed there, configured, how it interacts with each other, so that I would be ready to support it.
My knowledge is not enough to analyze the architecture, and I ask for advice, where to start the analysis, what can I read to gain knowledge? How can this be documented later?

I apologize for such a simple question.

Answer 1

Ahahah... At one time, I was left with several dozens of servers to support - on different OS - OpenBSD, FreeBSD, CentOS, Windows Server, and at that time I only picked Red Hat Linux a little ...
Simply because there was no one else leave the farm.
First of all, I made backups of all the configs that you can reach. Then he took the docks and tried to figure everything out on his own. Parse configs. There were no running through the forums. I just had to work a lot with my own head, which ultimately gave a considerable profit to the experience.

Answer 2

With a high degree of probability, the servers are used over the network, find out which applications are listening on which ports:
netstat -tulnp
Next, find the application configuration files that you see and study them.
In addition, see what you have included in autoload, how to do it, look for your version of Centos, the version can be found in the file /etc/centos-release
Yes, and do not forget to study the task scheduler
crontab -l

Answer 3

having received such servers, you need to check and back up and document a bunch of everything:
services, how access to these services is organized and managed, service configs, monitoring of these services, network and firewall settings, backup settings, and the most difficult thing is to look for custom crutches and automation - some things in cron or manually or through override corrected unit / init.d files, etc. and the more qualified the previous admin was, the deeper some custom things can be.

Answer 4

First of all, determine the purpose of the server - web, file, print, etc.
To do this, find the appropriate services there, you can simply enter from the root:

systemctl --type=service

And look. Further on the situation, if you find, for example, nginx - see its configs.

Answer 5

Start by backing up your servers. It's good if there is a full backup of the machine, especially if it is a virtual machine. If not, be sure to backup /etc, perhaps /opt/etc and /usr/local/etc (if they exist), there may be useful things in /usr/share and /usr/libexec, in /var (there are not configs, but status data, but selling them can also be sad).
See which services are up - depending on the version of CentOS, this is done differently, each service is its own config and a bunch of everything hung around.