My question is not really specific to Rocky Linux, but here goes.
On the latest count, I’m managing about a hundred installations: root servers, local servers, desktop PCs and laptops. Some of these are in our local school, some for a handful of companies all over France and even some in Belgium and Switzerland. Root servers are mostly Scaleway rentals in France. And then the inevitable desktops managed for friends and families.
Over the years I’ve been using various methods to keep track of all these installations, notably specifics like BIOS configuration details, exotic hardware, extra applications besides those I usually install, etc. In the beginning it was just a simple text file per machine, then I moved on to the CherryTree note-taking application, but I’m still not satisfied with the current solution.
Usually computer estate management software like GLPI or OSTicket can be relatively complex. I prefer to stick to the KISS (Keep It Simple Stupid) principle.
So I’m curious: what methods do you use to keep an overview over your installations ? In a nutshell, the sort of file or series of files where you can get all the relevant information about an installation in a quick glance ?
Thanks & cheers from France,
I’m pretty sure when I was using Uyuni with a load of connected hosts, it did have all the hardware info in it, etc, as well as a list of all the packages installed on those machines. It uses Saltstack to communicate, with the Uyuni server being the salt-master, and all the connected servers, etc, the salt-minion (client agent). Obviously it’s just for Linux, works fine for rpm-based as well as deb-based.
Uyuni is the open-source version of SUSE Manager, which used to be known as Spacewalk before SUSE took it over/forked it from Red Hat.
I had manual scripts until I encountered Ansible. Rather than “notes”, Ansible says “inventory” – a format of configuration data and package lists that it can deploy with “playbooks”.
Rocky 8.6 has package
ansible-core and RHEL documentation is apparently adding sections about “system roles”. They now document GUI, CLI, and Ansible way to manage RHEL.
Ah yeah, forgot about this, with ansible you can get the facts outputted to the console using debug module, which will also have the BIOS info, etc, etc:
- name: Get facts
- name: Get facts
running that playbook will then output to the screen all the facts, but you can easily dump that to a text file. Alternatively you can set hosts to
all and then all hosts will be dumped that you are managing with ansible. I expect if using AWX which is the open-source version of Ansible Tower, you could then have all of this graphically available like with Uyuni.
Ad hoc, without playbook you get the ansible_facts with:
ansible rocky -m setup
Where “rocky” is name of host(s). Could be “localhost”, “all”, group in inventory, or comma-separated list.
The setup module accepts parameters, for example to filter the output. See
Nice to know, haven’t used many adhoc commands to be honest.