Could also be a stale DNS cache entry on one device or the router. If you ping your duckdns fqdn from the device that can’t connect while on your home network, does it resolve to the correct public IP?

I still think a firewall/nat issue is more likely tho.

What is your router make and model? You need to enable hairpin NAT.

The port is forwarded from your router to the pi, right? If so, you could test for the router as the bottleneck using the router’s WAN side IP address as the target.

This should give you a good data point for comparison. If it’s also slow then you can focus on the router performance. Some are slow when doing hairpin NAT.

I’d say your chances are very good. Even their high end rackmount models work with the usbhid-ups driver. Don’t think they would change things up in this regard since they would also need to change their software.

It’s probably still IPv6 related. If you use something like Network Analyzer on your phone while only connected to the mobile network you may find that it only shows an IPv6 address and DNS server, no IPv4 config. That could explain the difference. Particularly if you were using the maximum typically permissible MTU. Your provider might also be doing some 6to4 tunneling somewhere that adds overhead and causes size problems.

You might want to do a DNS leak test from your phone with the wireguard connection down and then with it up to make sure you’re tunneling DNS. This will be clearer if you set pihole to use something upstream that an ISP is unlikely to use - quad9 for example.

Btw: does anybody know what bad things actually happen if there is no metal cage that blocks all the radio?

Noise happens. Could be no problem, or it could hurt your wifi or mobile data connections, or maybe raise a neighbor’s ham radio noise floor. I saw this recently when setting up a pi to run BirdNet-Pi. The USB3 connection to an SSD caused enough noise in the 2.4GHz band that the onboard wifi radio could only connect on the 5GHz band.

To start - moving services from bare metal to rootless Podman containers running via quadlets. It’s something I have had in mind for a while but keep second guessing the distro choice. Long-ish release cadence, systemd-networkd and a recent Podman version in the native repos, well supported, and not Ubuntu.

So far openSUSE Leap seems like the winner. A testing machine is up to install everything, write some deployment scripts, and decide on a storage layout and partitioning scheme.

If anyone has another distro to recommend that checks these boxes let me know!

I like rolling release for the desktop, but only want critical patches in any given month for this server, and a major upgrade no more than every 3-4 years. Or an immutable server distro. But it doesn’t seem like networkd is an option for the ones I’ve looked at (Fedora CoreOS, openSUSE MicroOS), and I am not sure if I want to figure out Ignition/Combustion right now.

Next project - VLANs on Mikrotik.

OP - Navepoint makes good racks for reasonable money. I have a Pro series 9u from them and it went together without any problems. It’s on the wall with a pretty big ups in it.

If you want to keep using networkd, you might want to consider if multiple interfaces are causing the wait. NM doesn’t care, but networkd gives more granular options for dependencies. If you have wired and wireless and only one in use the systemd-networkd-wait-online.service waits for a timeout period. You can find lots of info on it related to boot delays with that service.

Try the --any switch on the systemd-networkd-wait-online.service launch configuration. This will tell the wait-online service that any single routable interface is enough, you don’t need them all.

Run:

sudo systemctl edit systemd-networkd-wait-online.service

That adds the override.conf for the service. Add these lines:

[Service]
ExecStart=
ExecStart=/usr/lib/systemd/systemd-networkd-wait-online --any

The other possibility is if you have virtual .netdev devices configured (VPN, bridging, etc) and some of them are not essential for the machine to be online, you can set RequiredForOnline=no on the ones that aren’t essential.

It’s not bad to get running, and the alerting is really flexible. You can add Nagios and syslog alerts easily too.

Using it here. Love the flexibility and features.

I’m running NUT on the host os - no container. If that’s an option for you it will probably be much more reliable.

Huh. Losing USB access?

It’s not very exciting, but: Network UPS Tools (NUT).

Keep everything in good shape in the event of a power outage.

Here’s my messy-cabled 9u rack.

It has:

• Fiber gateway out of view on top of the rack.
• Switch, which also powers 2 Ruckus APs and 2 other switches.
• Mikrotik RB5009 router.
• Raspberry Pi x3 all running Debian Bookworm. I have too many pis right now, running Home Assistant, LibreNMS, Log collection, and a read-only NUT server that orchestrates shutdowns and startups on power loss. I need to consolidate these.
• 1L PCs. One is on Debian serving media and files. The other is a test server where I’m trying out Immich on openSUSE. I’m considering moving to that and rootless podman for services. To that end I have another of these 1L boxes on my desk trying other options (MicroOS, Fedora IoT, maybe others).
• HDs. These are backup drives for the 1L server. I keep them powered off except when needed.
• UPS and a managed, switched PDU.

Everything is set up for low energy consumption (~90w), remote admin, and recovery from power loss.

That and the shrinking ability to grant access to device storage. If that becomes an option only on rooted phones (which seems like the directly Google is heading) it will make the audience for such an app much smaller.