I have a (home) PVE node running for a few months and must say it is fantastic! (And, I'm so grateful that it's free.)
Yesterday I connected the UPS powering it, my gateway and NAS to this node via USB and installed nut... ever since then I've noticed I can't access the management webpage (port 8006) using Chrome (works better than Safari's private relay)...
My management device is a MBP on VLAN 1/Default... so same subnet as:
My PVE node (itself) is on VLAN 1/Default, uses a (Linux) bridged quad-port SFP+ NIC. The PVE node is only accessible via IPv4.
(All LXC/VMs are on VLAN 2, all with IPv4 & IPv6. All still running and their web interfaces accessible.)
I really don't see how installing (and half configuring) a single package could disrupt another package's web service, that's the only change I have made but I am looking elsewhere for the cause.
I can SSH into the node, all the containers/VMs are running normally. I suspect the node is running normally, except for the pveproxy service.
Using
(I don't know whether this is related to the problem, as I've never had to investigate things like this prior.)
The node's vmbr1 interface (vmbr0 is a pair of unused GbE NICs) has both an IPv4 (RFC 1918) address and an IPv6 LLA, no GUA.
I haven't setup any firewalling on this node (as I rely on my gateway), so I'm at a loss to explain why I can (a) get an IPv4 connection, and (b) it doesn't work in Chrome (it just says, "ERR_ADDRESS_UNREACHABLE").
Yesterday I connected the UPS powering it, my gateway and NAS to this node via USB and installed nut... ever since then I've noticed I can't access the management webpage (port 8006) using Chrome (works better than Safari's private relay)...
My management device is a MBP on VLAN 1/Default... so same subnet as:
My PVE node (itself) is on VLAN 1/Default, uses a (Linux) bridged quad-port SFP+ NIC. The PVE node is only accessible via IPv4.
(All LXC/VMs are on VLAN 2, all with IPv4 & IPv6. All still running and their web interfaces accessible.)
I really don't see how installing (and half configuring) a single package could disrupt another package's web service, that's the only change I have made but I am looking elsewhere for the cause.
I can SSH into the node, all the containers/VMs are running normally. I suspect the node is running normally, except for the pveproxy service.
Using
lsof -i -P -n | grep LISTEN I see it listening on IPv6 (only) port 8006, even though curl -I http://10.0.0.4:8006 (and curl -Ik https://10.0.0.4:8006 both) returns the expected output (HTTP 501) both locally and across the subnet from my management device.(I don't know whether this is related to the problem, as I've never had to investigate things like this prior.)
The node's vmbr1 interface (vmbr0 is a pair of unused GbE NICs) has both an IPv4 (RFC 1918) address and an IPv6 LLA, no GUA.
I haven't setup any firewalling on this node (as I rely on my gateway), so I'm at a loss to explain why I can (a) get an IPv4 connection, and (b) it doesn't work in Chrome (it just says, "ERR_ADDRESS_UNREACHABLE").
Last edited: