Logout after 30 sec or so



  • logs have:

    1. 12  webConfigurator authentication error for 'admin' from unknown WAN IP and 2  webConfigurator authentication error for 'user'  from same unknown IP.

    2. lots of  (connections.c.305) SSL: 1 error:1408A0C1:SSL routines:SSL3_GET_CLIENT_HELLO:no shared cipher

    3. Dec 9 20:30:51 logportalauth[71735]: Zone: lan - USER LOGIN: <>, xx:xx:xx:9e:20:e3, 10.10.0.166
      Dec 9 20:30:53 logportalauth[76726]: Zone: lan - Error during table lan creation. Error message: database is locked

    I tried restarting the webConfigurator, restarting radius, changing the admin password and setting the webgui session timeout to 0, 60 and 240 minutes.



  • It just fixed itself.

    Logs have 3 line saying " lighttpd[50578]: (connections.c.277) SSL: -1 5 54 Connection reset by peer". Releva?nt?



  • @pteek:

    Logs have 3 line saying " lighttpd[50578]: (connections.c.277) SSL: -1 5 54 Connection reset by peer". Releva?nt?

    Its says : "a client initiated a SSL (port 443) connection / transaction but I (lighttpd) can't make anything from it - I reset the connection".
    Solution: locate the 'client' - repair the device (setup) or forbid him to connect to your pfsense. done.

    Another know answer is: ignore these lines ;)



  • The logout is back. I can't stay logged-in info gui for more than 30 sec.



  • You login from where to where ?
    A device on pfsense-LAN ?


  • Banned

    Is this thread about webgui or captive portal?!?!



  • @doktornotor:

    Is this thread about webgui or captive portal?!?!

    At least these two ….. I guess.
    Featuring the issue "database is locked" ....



  • @Gertjan:

    You login from where to where ?
    A device on pfsense-LAN ?

    I login from a device on LAN. I am talking about the webgui.


  • Banned

    @pteek:

    I login from a device on LAN. I am talking about the webgui.

    So why are you posting captive portal logs? In general, CP should be on dedicated untrusted interfaces, not your trusted LAN.



  • @pteek:

    I login from a device on LAN. I am talking about the webgui.

    Well ….
    Session (ID) lost => cookie is not valid anymore => you get thrown out.
    This can happen because (just what pops up in my mind) :
    TCP sessions get destroyed because the line (cable) is bad, interface (check both sides) goes down, new LAN IP, switch in between is nasty .....

    No other logs to show us that might indicate bad connection ?

    Or, as seen often: browser (change for another one !) with really nasty plug-ins that messes up things, browser cache issues ....



  • That has no relation at all to the CP logs. Your browser's likely losing the session cookie somehow. Try the usual, clear cache, a diff browser, diff machine.

    Session cookie has no relation to the network-level connectivity, it's good even if source IPs change, new TCP connections, etc.



  • @doktornotor
    The captive portal also started throwing error 500 and 503.

    I just posted any error that I saw in the logs. Currently i can see

    lighttpd[52352]: (mod_fastcgi.c.1754) connect failed: No such file or directory on unix:/tmp/php-fastcgi-lan.socket-5
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.3021) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 0 load: 1
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.1754) connect failed: No such file or directory on unix:/tmp/php-fastcgi-lan.socket-4
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.3021) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 1 load: 1
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.1754) connect failed: No such file or directory on unix:/tmp/php-fastcgi-lan.socket-3
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.3021) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 2 load: 1
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.1754) connect failed: No such file or directory on unix:/tmp/php-fastcgi-lan.socket-2
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.3021) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 3 load: 1
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.1754) connect failed: No such file or directory on unix:/tmp/php-fastcgi-lan.socket-1
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.3021) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 4 load: 1
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.1754) connect failed: No such file or directory on unix:/tmp/php-fastcgi-lan.socket-0
    Dec 12 03:11:20 lighttpd[52352]: (mod_fastcgi.c.3021) backend died; we'll disable it for 1 seconds and send the request to another backend instead: reconnects: 5 load: 1

    @Gertjan

    I have tried different browsers, different machines. Their are no IP changes, cables have been replaced, no interface going down in the logs neither any complain from any user of internet disconnection.

    Current problems:

    webgui will logout after 30 or so seconds
    captive portal is throwing error 500 or 503



  • php-fastcti is dead for some reason, that's the source of the CP issues. The GUI doesn't use that though, something diff happening there. Could be a common cause making them both fail, like a dying drive might lose the ability to read a file (your session) making you log back in again, though usually those types of failures are severe enough that you won't be able to log back in at all as it'll fail to write the session.


Log in to reply