If you wonder what is a node with home internet exposure vs behind vpn would look like

[radicalrj]

I have 1.2TB of data, and I migrated my node from local internet exposure to a vpn proxy and the bandwidth dropping is considerable.

Also, it seems I am losing 0.01GB per day on this setup. It seems if I keep behind VPN, my node will lost all data in < 4 months.

Also, the VPN is really good (is actually a private VPN from a friends internet), the speedtest actually matches my internet connection.

After reading a lot, I guess this is the latency issue and somehow the vpn-node is not atractive anymore.

Hope this help others investing on node-vpn solutions. I don’t think it worth it.

[Vadim]

May be this VPN has lot of VPN nodes on its IP? have checked neighbors there?

[radicalrj]

no neighbors there, that is why i choose as vpn location and convinced my friend.

[LxdrJ]

What is the latency to your friends house?

[EasyRhino]

is that remaining bandwidth mostly ingress or egress?

I’m using a node that’s attached via a vpn to an oracle node and it gets the same bandwidth as everything else.

[arrogantrabbit]

@radicalrj

and the bandwidth dropping is considerable.

@radicalrj

After reading a lot, I guess this is the latency issue and somehow the vpn-node is not atractive anymore.

I would agree: I don’t see any performance drop using vpn on Oracle, that does not add any latency (maybe 2-3ms), the datacenter is very close to where I live.

@radicalrj

Hope this help others investing on node-vpn solutions. I don’t think it worth it.

How so? The choice is not between VPN or direct connection; the choice is between VPN or not running node.

I have two locations behind CGNAT, I run two nodes there over AirVPN. I don’t care about performance and I’m sure there are more nodes on those IPs — I did not check. Running nodes on those servers help me offset electric cost running them. Of course it’s worth it.

If you have public IP — why would you even consider VPN?

[dmp]

my node is behind nat so I have to use VPN. Before the “node success” settings, the filling did not at least stand still and was slowly happening. Now, as I understand it, filling is only happening in the USA. Here is my node that has been running since August.
PS. Anticipating questions, I’ll answer right away: No neighbors, good speeds, good disk, good computer. We were just unlucky with last year’s innovations (we were preparing to receive a bunch of data, so we “cranked up the settings” so that nodes were selected and tails were cut off) and we were unlucky with the location (Eastern Europe)