Newt failing to connect until restart "failed to get token"

### Describe the Bug

I have Pangolin running for a long while and a while ago my site start to show as "offline" in the panel, however, connections to it still working so I ignored it.

Today I updated Pangolin and everything else to latest version (newt was already latest before that) and was wondering if the update would fix the online status, I did not.

I then looked into newt logs and saw this:

I restarted newt and boom, it shows as online again. It just makes no sense to me that it was showing as offline for so long while the connection was working. And also makes no sense how it goes back to work with just a normal restart and no other fix from my side.

Seems like newt still has some issues with recovering healthy state (I had other issues with newt not recovering connections before, but was a different case and there was another issue for that in the past)

### Environment

- OS Type & Version: PVE > LXC > Docker (Newt) + Private VPS (Rackernerd) (Pangolin)
- Pangolin Version: 1.21.1
- Edition (Community or Enterprise): Community
- Gerbil Version: 1.4.3
- Traefik Version: 3.7.10
- Newt Version: 1.15.0
- Client Version: N/A

### To Reproduce

I am not sure how to reproduce this.

### Expected Behavior

Newt should be able to recover its health state, specially when it clearly can connect to pangolin server

Start with the Newt logs shown in the report and the listed Pangolin, Gerbil, Traefik, and Newt versions, focusing on the “failed to get token” state and what changes after a restart. Establish a reproducible failure or recovery scenario, then verify that Newt remains reported online and recovers its health state without a manual restart.

Written by the indexing model from the issue text.