Search results

Can the proxmox host ping 8.8.8.8? If you run apt update at the proxmox host cli what happens?

You say servers are exactly the same, but the server with the network down list ifb0 and ifb1 as available interfaces. Those are not listed on the working server with ip addr. Did you recently add a PCI card to this non-working server? Once you get the machine working again you really should...

have You tried the more straightforward method of using your vlan aware vmbr0 for your VM/LXC guest then assigning either vlan 2 or 3 at the vm-hardware- network device or LXC-network levels in the Proxmox webgui?

Agree and would go as far as to say the standard suggested FRR log level in the Proxmox wiki /etc/frr/frr.conf should be changed to warning as opposed to informational

Would also suggest a vm/lxc. In fact @tteckster has been kind enough to set up a github with proxmox scripts to easily startup LXC for basically everything you listed. Click the "server-networking" tab for tailscale and others. Pihole (my choice adguard) are under "adblocker-DNS"...

keep in mind that most would like a firewall between wan and Proxmox. Can be baremetal, your own rules in the Proxmox firewall or a vm/LXC.

Agree, something is broken with multicast over proxmox Linux bridges. I will test the couple VM/lxc currently needing multicast with OVS and report back. What does your ovs section look like from /etc/network/interfaces?

You could just install a VyOS, openwrt or OPNsens/pfSense as vm on Proxmox as well. Let that guest vm do pppoe and then have Proxmox access wan through the vm, if you don’t want to use additional baremetal hardware.

maybe your isp is blocking things when you try and use a default bridged config (the isp is seeing more than 1 MAC address at a single port) have you ever tried to use routed or masquerading network config for internet access? Also curious what DNS your isp provides through DHCP?

semantics of the wording “pass through” might make things murky for those helping. I infer you mean “pass through“ in the generic non-networking way and not IOMMU/SR-IOV. is this external modem a firewall/router. If not any consideration for exposing the Proxmox host directly to wan? Is Proxmox...

I would slow down, your /etc/pve/corosync.conf might already be messy. I have a separate redundant corosync network, but I think like most home users that go this far I first set up clustering through the GUI using my vmbr0 network/IP on all nodes. Then I later added the separate Corosync...

This looks like corosync expects the nodes to be using 10.10.10.1 and 10.10.10.2 not the 10.10.1.10/24 10.10.1.20/24 and 10.10.1.30/24 you set up earlier today.

Correct and Congrats

Since you are starting with a separate cluster network prior to having a cluster try the link below. Read here, https://pve.proxmox.com/pve-docs/pve-admin-guide.html#pvecm_redundancy , find the section copied below to start. Then follow the link within that section to add to a separate cluster...

So you can do either: 1: IOMMU passthrough and select the PCI device directly from the OPNsense vm network hardware setup. Then Proxmox has no way to see either nic. 2. No IOMMU, but use the Linux bridge directly as the OPNsense wan and lan. Both Proxmox and OPNsense have access to the nic.

Read the last line of @mira post above and follow the directions from that link on how to edit. This link will also tell how to work around that fact that edits take effect immediately. As he stated don't forget to increment the config_version (currently at 4 above, increase to 5) when you save...

Neither enp1s0 or enp3s0 are showing up with ip a, so they are not available to use. just in case, did you attempt to do iommu passthrough of the pci card to opnsense? Can't do that and use the same nic as the port for a Linux bridge. if not, maybe consider trying the newest 6.1 kernel, pinnED...

This sounds like a setup issue with pfSense letting vlan5 traffic to/from wan then Or DNS. Can the vm ping 1.1.1.1 or 8.8.8.8?

Look at the bottom of you /etc/corosync/corosync.conf file. Even the one posted above find the totem section, then under that the interface subsection for for your link0 and link1 lines. add the knet_link_priority: lined as @mira stated and Corosync should now use your private ring1 as primary.

Ok, have You double checked the managed switch ports? Not sure of your switch manufacturer nomenclature, but is the switch port connected to eno4 untagged vlan5 and is the switch port connected to pfSense tagged for vlan5? is the Proxmox firewall unchecked on the vm network interface? for...