A lemm.ee refugee ;)

  • 18 Posts
  • 291 Comments
Joined 1 year ago
cake
Cake day: June 6th, 2025

help-circle






  • The 2008 crisis and how the government reacted is partly why I live frugally

    I understand this sentiment. As someone who didn’t see wealth during their childhood, people get confused as I still travel by public transport instead of driving. I just assume that things can go sideways at any moment. Yes, it can seem like I’m a bit paranoid, but it’s not illogical.

    Thanks for sharing.












  • I think you have a typo here, eth1 is WAN not WLAN

    Corrected!

    When the VM goes down you can still configure the network manually on your PC and access proxmox directly

    Yes, that’s why having this MGMT port is so important. Time to make some more cables :P

    This is why the hardcoded ip will still work.

    For some of the critical infra, I will probably hardcode the IP

    I saw that you added the detail that you’re also running another vm - AFAIR you could route that via the management port as well with via proxmox without added complexity.

    I’ve elaborated upon this in this comment in this thread. Please check and let me know your opinions.


  • I have one more doubt.

    (Nothing setup as of now, just checking all possible angles)

    Since, as per your suggestion, I have passed the NIC to OPNsense, so Proxmox won’t see the dual NIC card. The onboard ethernet port eth3 is connected to a switch and Proxmox will use eth3.

    Here’s a fresh diagram

    You can see that I have a Ubuntu VM. How will the PC communicate with the VM? (Links E and D are hypothetical. I don’t know which one is better given my scenario). Will it be C -> B -> D or C -> A -> E?

    Case #1: C -> B -> D

    The switch must differentiate between uplink traffic/proxmox traffic/other physical devices connected to the switch (phone/TV). I prefer this solution because even if OPNsense is down, I can still access the Ubuntu server. But the switch must handle this complexity. I will be getting a managed switch anyway.

    Case #2: C -> A -> E

    The interface eth2, which OPNsense is using, must have some sort of routing table which redirects to links E or F depending upon destination IP (Probably implemented using virtual bridge or virtual switch). This is simpler for the switch but if OPNsense is down, i lose direct network access to the VM. I can probably access it via Proxmox web-interface.

    Can you tell me which one is preferred or which one you would have done?

    Sorry for my ramblings.



  • Have done something similar, you’ll discover the joy of have no internet while you do server maintenance, upgrades or disaster recovery and worst case, hardware failures.

    I may purchase another lower powered device to prevent a single point of failure, but I want to get this thing working. Sorry, too excited with Proxmox now :D Also, in the worst case scenario, I’ll just use my current router while I manage my OPNsense