Hello,
We’re seeing intermittent outbound IPv4 connectivity issues from machine `e8272d1b7d77d8` in `fra`.
This started around April 14, 2026. The setup had been working reliably for about 5 months before that, and we had not changed the configuration.
To rule out our setup, we rebuilt the container as plain Alpine with no Tailscale logic and tested again from inside `fly ssh console`. The issue still reproduces.
Summary:
- DNS looks healthy: repeated `nslookup` tests for multiple domains were `10/10 ok`
- Raw IPv4 HTTPS is unstable even without DNS:
- `https://1.1.1.1`: 1/10 ok
- `https://8.8.8.8`: 3/10 ok
- `https://9.9.9.9`: 2/10 ok
- Public HTTPS endpoints also intermittently time out from inside the machine
- Failures are `curl: (28) Connection timed out after 5002 milliseconds`
Please see the full log at: gist:04d0ee4de9d6267c8b3585c5ed2758bc · GitHub
Since this reproduces on plain Alpine, against raw IPv4 addresses, and from inside the machine itself, this appears to be a machine/host/regional outbound networking issue rather than DNS, Tailscale, or app configuration.
Could you please investigate outbound connectivity for this machine in `fra`?
Machine: `e8272d1b7d77d8`
Thanks.