As the title says, I want to know the most paranoid security measures you’ve implemented in your homelab. I can think of SDN solutions with firewalls covering every interface, ACLs, locked-down/hardened OSes etc but not much beyond that. I’m wondering how deep this paranoia can go (and maybe even go down my own route too!).

Thanks!

  • Pika@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    27
    ·
    edit-2
    9 months ago

    My security is fairly simplistic but I’m happy with it

    • software protection

      • fail2ban with low warning hold
      • cert based login for ssh (no password Auth)
      • Honeypot on all common port numbers, which if pinged leads to a permanent IP ban
      • drop all firewall
      • PSAD for intrusion/scanning protection (so many Russian scanners… lol)
      • wireguard for VPN to access local virtual machines and resources
      • external VPN with nordVPN for secure containers (yes I know nord is questionable I plan to swap when my sub runs out)
    • physical protection

      • luksCrypt on the sensitive Data/program Drive ( I know there’s some security concerns with luksCrypt bite me)
      • grub and bios locked with password
      • UPS set to auto notify on power outage
      • router with keep alive warning system that pings my phone if the lab goes offline and provides fallback dns
    • things I’ve thought about:

      • a mock recovery partition entry that will nuke the Luks headers on entry (to prevent potential exploit getting through grub)
      • removing super user access completely outside of local user access
      • Pika@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        3
        ·
        edit-2
        9 months ago

        I just expanded the existing fail2ban config on the commonly used default ports such as 22, 21 Etc, any requests on those ports get sent into purgatory, so the ip gets blacklisted any connections from it hangs until it times out. It’s a super basic setup iptables logs whenever a request is not in the current firewall (last rule in the chain) and then fail2ban reads the log and handles the block. I don’t count it as part of the normal setup because they’re isolated Because the actual ports the service is on still have the normal rule set but the default port numbers are just an instant if there’s activity on it you’re gone