Messages

# opnsense-revert opnsense


Cheers,
Franco

You installed the FreeBSD package manager version, which isn't good.

>   opnsense-26.1.3 [OPNsense] (Vital flag changed: 'true' -> 'false')

This looks like a weird database bug. Vital flag is never removed.

Try this first...

# opnsense-revert pkg


Cheers,
Franco

> On pf long ago, Suricata was multithreaded which gave a performance boost over Snort.

The key part here is Netmap+Suricata.


Cheers,
Franco

Register rules for DNAT are automatic rules and only visible with inspect in the new GUI, yes. That's also the reason you can't edit them directly.


Cheers,
Franco

Thanks, I've queued this up for 26.1.4.


Cheers,
Franco

Disk full perhaps?


Cheers,
Franco

Already reached out to him.


Cheers,
Franco

To make sure it's not the plugin try

# opnsense-revert -r 26.1.2 os-upnp


Cheers,
Franco

https://github.com/opnsense/changelog/blob/7c2bcc9c4b7f6f4858a6105ff531536f4881297d/community/26.1/26.1#L123

https://docs.opnsense.org/manual/firewall.html#rules-new

> I guess the old rules are active until you delete them and then the new ones take over?

Both can and will be used.


Cheers,
Franco

Run the command in a root shell to apply it. It's easy as that :)


Cheers,
Franco

I wrote a small haiku-ish poem for you.

Complaints, here before
Complaints, now 20-20
forever fruits of work


Cheers,
Franco

> If I wanted to reverse it and put the RRD data back, would it be:

No, it creates a backup when you allow it to flush, which it also says...

Sounds like it works as expected now.  :)


Cheers,
Franco

My theory is if you do an "opnsense-bootstrap" the error will persist through a complete reinstall, which means it's a local issue with some stray file left for whatever reason and not a global one.

It's worth a try but chances are it doesn't help.

Just as an additional data point:

> community/24.7/24.7:OpenVPN DCO support, FreeBSD 14.1, Python 3.11 plus much more.

So we've had 3.11 for 1,5 years in total which roughly matches your "24" and "25" were working.

Cheers,
Franco

> Correct, but on a fresh 26.1.2_5 upgraded to 26.1.3 exist those files as well:

Yep, but they don't error because filter.rules is populated with legacy rules by default.  :)

We have enough examples elsewhere and don't reference these files so removing seemed like the obvious choice here. Will ship in 26.1.4.


Cheers,
Franco

At this point it seems that this is a more fundamental OS/Python issue than update related. We've found forum posts from the past regarding similar issues. Not a lot of clues on why this happened, but they all seem to have tainted Python load orders.

Did you install anything via pip?  Just trying to find out what's wrong that the audits aren't telling us.


Cheers,
Franco