Error "Fatal error: Uncaught Error: Call to undefined function gettext() in /usr/local/www/index.php:34 Stack trace: #0 {main} thrown in /usr/local/www/index.php on line 34"

emil

@eli This is exactly what we are experiencing here almost on all pfSense firewalls. After firewall or PHP-FPM restart error is back like in 10 minutes.

emil

I have ran commands and restarted PHP-FPM in console option 16. All is working good so far.

pkg update -f

pkg upgrade -f pfSense-repo

pkg info | grep pfSense-repo

at this moment make sure it's running version 2.4.4_7 of this package

/etc/rc.php_ini_setup

stephenw10

What pfSense packages did you have installed anyone who hit this?

Anything unusual?

Steve

emil

Just Cron.

rbgarga

@emil Do you have any extra item on your crontab?

emil

@rbgarga

- - - - root /usr/local/bin/php /usr/local/www/pfmonitor.checkin.php

stephenw10

You have pfmonitor installed then?

Anyone else hitting this running that?

Steve

emil

@stephenw10 Yes, I have pfmonitor installed.

rbgarga

@emil I don't know the code of that pfmonitor.checkin.php, but I strongly believe it was the cause of the issue. I'm saying that because we tried several times to reproduce the issue here and we couldn't. That is the reason we don't support systems running 3rd party code.

emil

@rbgarga Yes I do understand, I have contacted PFmonitor support also to report the issue. Thank you for your help.

emil

Here is a response from PFmonitor support I have received:

"12-17-2019 ALERT! Firewalls Showing offline, PHP Uncaught Error

Last night or early this morning, a problem emerged suddenly with firewalls web interfaces going offline, and showing an error message of "Fatal error: Uncaught Error: Call to undefined function gettext() in /usr/local/www/index.php:34 Stack trace: #0 {main} thrown in /usr/local/www/index.php on line 34"

From what we can tell, there was some sort of repository update or change last night on pfsense, and as this change filtered down to the firewalls, it caused an issue with the agent. We are still investigating what was changed, and why this occurred. But
as a precaution to having any more units get stuck, we have issued the uninstall command for all agents, syslog is un-affected by this. We would rather drop off the agents, than have any more firewalls go offline.

If a firewall is already offline there are two potential ways to fix it:

A simple power cycle should allow the agent to uninstall itself during the reboot in most cases.
From the shell on an affected unit, run this command, and then reboot the unit: rm pfmonitor.checkin.php: "

MichaelBoye

Hi.
We have the same issue on at least two of our customers firewalls.
Both of them run 2.4.4 and is in pfmonitor.
I tried factory default one of them yesterday, but that didn't fix the issue.

I will try and run the "rm pfmonitor.checkin.php".

BR,
Michael Boye

eli

This script did it for us via the shell:

pkg update -f
pkg upgrade -f pfSense-repo
pkg info | grep pfSense-repo
-Verify version 2.4.4_7
exit
option #16 to restart php-fpm