What network admin tools are actually worth learning in 2025?

[woodchucksage]

Hey STH crew — I’m revisiting my home lab/network stack and want to sharpen my sysadmin toolkit.

I know the classics like:

pfSense / OPNsense

Ansible

NetBox

Zabbix

Nmap

iperf, Wireshark, etc.

But I’m curious what others are really relying on day-to-day. Not just flashy GUIs — tools that save your ass during outages, help with config automation, multi-node patching, or make documentation less of a nightmare.

What tools do you trust to run clean infrastructure at scale?

Bonus points for ones that:

Work well in mixed Windows + Linux environments

Let you document and re-use scripts

Offer agentless automation

Or even tie into Grafana, Prometheus, or webhook workflows

Let’s build a 2025 toolbox list

 

[mrpops2ko]

eBPF and consuming as much as you can from brendan gregg on this topic

your question is very broad and im not fully sure what you are asking, is it an SRE question? because generally that comes in 2 parts - fighting the fire (going from a non-working state to a working one) and then all the analysis / understanding / monitoring / post diagnosis stuff afterwards

a lot of the concepts i steal from other disciplines because it seems like software development has become so all encompassing it seeks to take over all other industries, meaning that the whole idea of unit tests and constant monitoring are the kind of thing that you might want to consider

so for example having scripts from various locations all trying to do specific things that have a specific outcome and to ping you on the outcome not being what you want (say being able to bypass a vlan or route somewhere where it shouldn't) because the larger and more complex your network(s) get the harder it is to understand, know and even monitor when a change is done whether it has some kind of knock on implication that wasn't known / considered

i remember reading a really interesting article about a guy who managed to get into the entire comcast client list from an api because one of the nodes had some weird routing internally and an internal > external route managed to exist - my guess was that they were all separate and then someone decided to throw them into a load balancer... let me google it

ok after 20 minutes i cant find the article anymore, which kind of sucks to know things just seem to get purged from the internet... nevermind i found it!! well worth a read

 

[CyklonDX]

sometimes shell/bash/python scripts (or with other scrappers) to grab data from switches/routers and feed them into influx/opensearch. *you can log in events, stats from time to time, and so on...

netcat is nice for digging deeper.

 

[i386]

Git, Subversion for versioning config files, scripts and other things
Passwordstate, bitwareden, keepass(xc) or other password managers
Jira, GLPI or other ticketing/support desk systems
Confluence, mediawiki or other wiki system
Sharepoint ( )
GLPI or other asset/inventory management system
an editor that can open and handle large (think 4GB+ log) files
Excel, libreoffice calc (I abuse excel to manage/plan my network at home)
winditstat, kdirstat or treesize
visual studio code or other lightweight ide that makes it easier to write small application that make life for admins easier

Edit: veam of course (or other backup software)

 

[chemisesociety]

Here’s what’s been saving me time lately:

• NetBox + Grafana for mapping and visuals
  
  
• PowerShell DSC for Windows config, alongside Ansible for Linux
  
  
• Zabbix with webhook triggers for alerts
  
  
• Attune runs Bash/PowerShell/SQL scripts across multiple nodes without agents.
  
  

 

[Scott Laird]

I'm kind of fond of Clickhouse as a database for log-like things, and Vector for munging logs on the way to Clickhouse. Grafana is perfectly happy talking to it, and it's disturbingly fast with good documentation.

I kind of hate Elasticsearch/Opensearch for as a log store. Or pretty much anything else.

 

[unwind-protect]

I second the eBPF recommendation. Now there is a near universal tool.