Performance drop after routing through Pangolin

Third Party Help
1

Hi everyone,

this is my first topic here - was a silent reader until now. Reading here was very helpful in the past. I took a look at the search-function but did not find anything related.

I’m hosting Vaultwarden in my homelab inside a Proxmox VM. I also run a Nginx Proxy Manager instance there on another VM. Recently, I added a Pangolin instance on a remote vServer to my setup to improve my privacy and security.

Previously, my domain was connected directly to my home network via DynDNS, going straight to the Nginx Proxy Manager on my local setup. Everything was fast and responsive.

Now, the setup looks like this:

  • Domain → A record → Pangolin

  • Pangolin → newt tunnel → Nginx Proxy Manager

  • NPM → internal → Vaultwarden VM

Everything works perfectly over HTTPS, end-to-end. However, since introducing Pangolin, I’ve noticed a significant performance drop - the web-interface and client-apps take noticeably longer to load and sync. But no problem with functionality

Looking into the logfiles was not helpful. I found no errors at all.

The issue doesn’t appear to be related to resource limitations: the vServer and my homelab both have plenty of capacity, and other services are still lightning fast.

Has anyone experienced similar slowdowns when routing traffic through Pangolin? Any suggestions on where to start troubleshooting or optimizing this setup would be greatly appreciated.

Thanks in advance!

2

I’m running Pangolin on a cloud VPS and using Newt in my homelab to connect my local services. NOWHERE in this scenario does NPM figure in at all. If this is your scenario, I think you’re doing it wrong.