Messages

Thank you both for the clarity. I wasn't sure so I didn't lead with a "sky is falling" reaction to the find.

I just noticed that the firewall widget for the dashboard of 26.7.2 indicates two sections of the pie graphic as 'let out' and none as 'let in' which seems odd to me.

https://www.pasteboard.co/WK01WqcDhUFU.png
 Image vanishes in November.

IMHO, since logs could also be inaccurately tagged by a clock being set improperly, using the OPNsense device as the clock reference would allow a check against that if no other view of clock value is used in the gui.  Perhaps a config toggle?

[privacy.resistFingerprinting]

Sorry, my prior post edited while you quickly responded to my initial version. :/

The inaccurate time corresponds to UTC time.


It is indeed <string of explicatives> Firefox causing 100% of the problem.  The about:config setting privacy.resistFingerprinting when enabled forces time to be UTC.

Code Select Expand

# sysctl machdep.disable_rtc_set
machdep.disable_rtc_set: 0

My search for cmos clock ignore was within OPNsense so I failed to discover what FreeBSD may provide.  Trying to set that sysctl at 1 and see if this solves.. no luck, though the sysctl now reports as 1, the traffic graph is still not in sync with the other clock. 

Here is the WEIRD thing, I just noticed that the timestamp for the most recent firefox screenshot is also wrong, it has the same time as the traffic graph when taken. dclock on my system (primary box) shows correct and the sysctl is not set to ignore cmos. On both boxes, date "+%Z" reports CDT. 

Are we seeing yet another damned firefox-does-it-by-itself feature causing unintended problems?  I do not see anything in firefox about:config that conclusively relates to clock or time or timezone.

It appears that the traffic graph does not use the OPNsense system clock but perhaps pulls the time from the motherboard by some other method?  I say this because I recently went to the bios of the motherboard to discover it showing the same incorrect time as the traffic graph at that moment, the minutes were correct and even the date, but the clock had to be reset to 10am from 3pm in my bios.  I suspect that the cmos battery is no good, as the traffic graph continues to use an inaccurate time. ntpd works properly or the correct time would not be displayed by the system info widget. I hoped to find any config option to ignore the cmos clock but could not find anything about it while the ntpd/system time is accurate. Attached is a screenshot that illustrates the disparity.

I am very pleased to report that my update from 19.7.x to 20.1 worked exactly as expected.  Either my custom config tunable vm.pmap.pti was retained or OPNsense now handles it better for my processor.

To all those involved, thank you very much!

It is an AMD Phenom(tm) II X4 960T.  I believe I have had the same issue with an AMD Athlon II X2 270 Regor.  However, so long as I am able to recall the sysctl to adjust, I can progress past the bump in the road and get an update just fine.

[system > settings > tunables]

Hello!

Possibly I am in the minority of users who have AMD hardware, so when an update fails causing continuous reboot I may eventually remember that I need to solve it as below:

choose boot loader
set vm.pmap.pti=0
boot

After bootup, go to system > settings > tunables to re-create the line
tunable: vm.pmap.pti
description: annoying AMD unfriendly mitigation
value: 0

IMHO, it would be helpful if this mitigation did not cause trouble for processors other than intel, could processor be detected and sysctl automatically inserted as above? At the moment for me it is simply remembering the solution.

[Feature req:]

Error "This Hostname, IP, MAC address or Client identifier already exists"
The error message implies that if any of those is duplicated, setting the static lease is prevented. 

• If a box has both ethernet and wifi, and you try to set a static lease for both forms of connectivity, it balks when the second one has an identical hostname. Any box is not limited to a single ethernet port/card, nor wifi, and neither necessarily exclusive, so shouldn't it be possible to have multiple forms of connectivity?  I believe it should issue a warning note that a duplicate hostname was attempted, with a confirmation prompt whether to keep the duplication.


• Feature req: (given the above is resolved to permit duplicate hostname)
  If the intent in the future is to reach a specific host through an IP, yes uniqueness is a good thing, but could it seek out an alternate IP/connection from those of the same hostname when one IP fails?
  

(on OPNsense 19.1.b_492-amd64)

Hello!

During my recent install & setup of OPNsense, the following crash report was generated, but since I also had not perfected internet connectivity by then, it could not be sent.  Perhaps such a generated report could be held somewhere when unconnected and then sent by choice or automatic?  I hope this report helps, and yes I have seen the same or very similar issue/error on pfsense.  Either some twiddle of the config or a reboot or both seems to have cleared the problem for me.

Code Select Expand

amd64                                                                                                                                                                                             
10.1-RELEASE-p6                                                                                                                                                                                   
FreeBSD 10.1-RELEASE-p6 #0 5aa5ada(master): Thu Feb 26 16:26:03 CET 2015     root@sensey64:/usr/obj/usr/src/sys/SMP                                                                               
                                                                                                                                                                                                 
Crash report details:                                                                                                                                                                             
                                                                                                                                                                                                 
PHP Errors:                                                                                                                                                                                       
[05-Mar-2015 00:18:36 America/Chicago] PHP Fatal error:  Uncaught exception 'Exception' with message 'Timeout (60) executing :service reload all' in /usr/local/opnsense/mvc/app/library/OPNsense/Core/Backend.php:70                                                                                                                                                                               
Stack trace:                                                                                                                                                                                     
#0 /usr/local/etc/inc/util.inc(162): OPNsense\Core\Backend->sendEvent('service reload ...')                                                                                                       
#1 /usr/local/etc/inc/pfsense-utils.inc(834): send_event('service reload ...')                                                                                                                   
#2 /usr/local/www/wizard.php(901) : eval()'d code(33): reload_all()                                                                                                                               
#3 /usr/local/www/wizard.php(901): eval()                                                                                                                                                         
#4 {main}                                                                                                                                                                                         
  thrown in /usr/local/opnsense/mvc/app/library/OPNsense/Core/Backend.php on line 70                                                                                                             
[05-Mar-2015 00:19:06 America/Chicago] PHP Fatal error:  Uncaught exception 'Exception' with message 'Timeout (60) executing :service reload all' in /usr/local/opnsense/mvc/app/library/OPNsense/Core/Backend.php:70                                                                                                                                                                               
Stack trace:                                                                                                                                                                                     
#0 /usr/local/etc/inc/util.inc(162): OPNsense\Core\Backend->sendEvent('service reload ...')                                                                                                       
#1 /usr/local/etc/inc/pfsense-utils.inc(834): send_event('service reload ...')                                                                                                                   
#2 /usr/local/www/wizard.php(901) : eval()'d code(33): reload_all()
#3 /usr/local/www/wizard.php(901): eval()
#4 {main}
  thrown in /usr/local/opnsense/mvc/app/library/OPNsense/Core/Backend.php on line 70