Messages

Thank you! I guess I was looking at the first tab.

I installed the plugin per the instructions here:  https://docs.opnsense.org/manual/how-tos/nut.html

The options looks somewhat different, but notably the username/password fields are missing so I'm unable to connect to the server that has the UPS on it.

Can I assume configuring from the command line is the same at other NUT installs for this or will the issue be corrected in a future release?  It is somewhat unusable as a client right now unless I'm missing something.

Same problem for me with the Traffic Graph. Latest MS Edge on Windows 11 and macOS 14.6.1.

All were/are fine, link lights and all.  First thing I did when things went wrong, actually.

I've been using OpnSense for as long as I can remember. I have switched hardware a number of times. Today, I ran into something I've never seen and just can't figure out.

My current hardware has 3 interfaces: em0, em1, and em2.  Only em0 (LAN) and em1 (WAN) are used.

My drive has been having some issues reported by SMART, but otherwise everything has been working.  Today I decided to backup my configuration, put a new drive in and restore from backup.  Nothing I have not done before.

No WAN internet. After 2 hours combing over my config, on a lark I decided to reassign em2 as WAN and unassign em1.

Everything works perfectly with that change. I did nothing else. I have not tried swapping them back yet.

I have Unbound DNS configured as my DNS. I do use DHCP also with some static IPs defined to hosts. I have some LAN hostnames defined and it all seems to work okay... except when the WAN goes down (like it is now, I'm on my phone's hot spot).

I'd think that local hosts should still be able to resolve to each other. I must be missing a setting or maybe a rule to allow this.  Is this enough information to get a few suggestions?

Thank you! That seems to have done it.

I'm getting an error in System -> Settings -> Administration.

Even though I have only http selected, I'm getting an error when I save: "Certificate webConfigurator default is not intended for server use."

This cert does exist in the https dropdown, but I'm not using https.

I installed a fresh copy on a test machine and there appears to be a different default TLS cert on a clean install which is NOT in my production install.  I have upgraded many, many times so perhaps that is part of it.

Obviously, with this error, I cannot Save any changes in that section.

Well, I got things a bit better after I discovered the NAT Reflection options in Advanced.

Ticking the Reflection for Port Forwards and Automatic Outbound Reflection got me to the server.

Unfortunately, Apache's DNS site detection is broken because of this on the LAN.  www.mysecretsite.com resolves as the "first" site, but www.myREALLYsecretsite.com resolves to the first.

What I really need to understand is why Unbound's overrides are not working for this but are for other items.  If my LAN clients are hitting it, shouldn't its overrides be honored?

All, I have an internal DNS server that I want to retire in favor of using the built in UnboundDNS.  Everything works except my web server behind NAT.

Externally www.mysecretdomain.com resolves and works perfectly.

Internally www.mysecretdomain.com throws a rebind error because it tries to go to the admin page of OpnSense instead of www which is on a different system.

Indeed, internally pinging www.mysecretdomain.com resolves to my PUBLIC IP when it should resolve to my INTERNAL IP.

Even going to the Overrides section of Unbound and making sure www.mysecretdomain.com resolves to 10.0.1.201 does nothing.  It insists on resolving to my Public/Opnsense WAN IP.

What the heck am I doing wrong?

A long time ago when I started running OpnSense it was advised not to use SSDs.

Has this changed?  I have a mini PC with an EVO Plus NVMe drive I can upgrade with.

[*solved*]

*solved*

Over in the FreeBSD bug tracker I found this link, and see comment #48.

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=230172

I have to add this to /boot/loader.conf.local on my systems - I figured I try it on the Dell (which I am using) and the Apple which had the same issue.  It fixed both systems.

kern.vty="vt"
hw.vga.acpi_ignore_no_vga="1"

Thanks for responding. I did download vga.  I've sense tried it on a Dell 3050 also with EFI firmware.... same thing.

Did I download the wrong image?

I was looking in Tuneables as someone in the FreeBSD forum thought disabling vesa would do it.

Okay, I went back to this with fresh eyes.  Maybe someone has an idea about this:

Only the local console on the display is "hanging". OPNSense is alive and well both via the web and ssh ports. It works.

I was used to seeing information on my old setup after the loader, so it must be something in that loader.conf for video drivers?

I don't know how much I even need that since I run it headless, but it was nice to have a working monitor with it as a diagnostic tool as OPNsense loads/shutsdown if there is a problem.

Thoughts?

This is a culmination of my two other posts. My simple morning project has turned into a nightmare. That's how it goes for me.

I decided to switch out my OpnSense hardware. I have a decent i5 Mac mini not doing much, so I download the image, got it installed and my configuration imported. The initial boot after the installer was successful and everything came up and I'm ready to swap. Awesome!

I shut down to move it into it's new home and it's hanging at /boot/defaults/loader.conf

I reloaded, was successful again, ...and again... shutdown and first cold boot, it hangs.

I don't get it and I don't know enough about FreeBSD to even begin to figure out why this seems to work fine then fails.

Um, help?