Fail2Ban VS fwknop

now some things you need to think about: - cloud init - this will need to be secure so lock it down hard anything not needed an alternative OS to look at if you have the ability's is https://www.alpinelinux.org/ also as these devices are not that powerfull every extra agent / abstaction layer you add impacts performance need to look at low over head security https://www.crowdsec.net/ and https://github.com/fail2ban/fail2ban (if you call fail2ban security) - using certificates to authenticate ssh login

See https://github.com/fail2ban/fail2ban for beginner's guide, basically you set it up to monitor logfiles and it would act accordingly (plenty of built-in config to handle various daemons so you don't have to write yourself).

- Nginx & crowdsec/fail2ban if you are exposing your parts (services) to the public ( https://hub.docker.com/r/baudneo/nginx-proxy-manager, https://www.crowdsec.net, https://www.fail2ban.org )

— In /etc/fail2ban/action.d/cloudflare.conf I copied the file from https://github.com/fail2ban/fail2ban/blob/master/config/action.d/cloudflare.confand added my ‘cftoken’ and ‘cfuser’ on the bottom

https://github.com/fail2ban/fail2ban is a mature, easy to set up way to have some dynamic firewall rules that respond to attacks. There are more sophisticated options, but they are probably not worth the return on time investment for you.

You can create your own regexes for custom services: https://github.com/fail2ban/fail2ban/wiki/Developing-Regex-in-Fail2ban

Others seem to be (or were) experiencing this too: https://github.com/fail2ban/fail2ban/issues/3100

sounds fun; i see the arch aur has a few options as well. have you tried https://www.cipherdyne.org/fwknop/ ?

Yes that's the basic idea, i tried to use fwknop first but it didn't do what i wanted it to do so i made my own

Port knocking. Or better yet FWKNOP. I'm disappointed I don't hear people talk about it more. The port isn't even open until you give the secret combination of knocks on a large number of ports. There's much more to it. I recommend listening to Episode 865 ofSecurity Now.

> Is this approach used elsewhere?

Yes, or at least in a similar fashion. An alternative variant of port knocking is SPA (Single Packet Authorization). Often SPA protocols use UDP and contain within the body field an encrypted payload containing all the required data to authenticate and authorize a particular request.

There are multiple different implementations of SPA: OpenSPA [1] (full disclosure: I am the author of OpenSPA), fwknop [2] just to name a few.

SDP (Software Defined Perimeter) often builds upon SPA technologies in order to achieve a form of zero trust access.

[1] - https://github.com/greenstatic/openspa

[2] - https://github.com/mrash/fwknop

I am currently re-writting the OpenSPA protocol (version 2) and I plan on playing around with eBPF as well, so thanks eeriedusk for paving the way :)

As an alternative to port knocking, there is: https://github.com/mrash/fwknop

fwknop is nice and simple

Sure, few links for when you dig in: http://iplists.firehol.org/, https://crowdsec.net/, https://www.zeroflux.org/projects/knock/, https://www.cipherdyne.org/fwknop/

An upgrade to port knocking is Single Packet Authorization [1]. It doesn’t suffer from the observability, and other, problems of port knocking.

[1] https://www.cipherdyne.org/fwknop/

Or keep the port closed like I do with my ssh port and use fwknop to open the port only when needed.