Captive Portal is broken?

[GeorgeCZ58]

@damiengombaultrecia Hi, the same happen on my sites. How to troubleshoot this?

[Gertjan]

@georgecz58 said in Captive Portal is broken?:

troubleshoot

It exists. Here it is : Troubleshooting Captive Portal.

Start by installing the System_Patches pfSense package 2.0_6. It permits you to apply changes that are made avaible without having to wait for the next release.
Some patches are already prospoed in the System_Patches upon upgrade of this package. Other can be obtained as they are created (source : this forum, redmine or Github).

Take note : as of today, September 1st, there are no available patches for 22.05.

As said already above, I've added one patch.
Read about here https://redmine.pfsense.org/issues/13323
This is the patch I used as a patch ID : https://github.com/pfsense/pfsense/commit/add6447b9dc801144141bb24f8c264e03a0e7cae.patch

This patch is useful when you have to add 'bypass' MACs.
I'm not using those right now on my portal.

True, some functionality doesn't work well or not at all.

See here : FreeRadius and quotas, doesn't work since 22.05 and if you use multiple portals (thus multiple interfaces) : Problem with multiple Interfaces since Version 22.05
For the last issue, there is a redmine present - and isn't even a patch proposed : just edit the file ( add a dot somewhere ;) )

I know it's easy to say that the captive portal doesn't work.
Keep in mind it's also easy to say it works. Just have a look here : https://www.test-domaine.fr/munin/brit-hotel-fumel.net/pfsense.brit-hotel-fumel.net/index.html#portalusers

The connected users are not my kids, neither my family, neither my collogues. I even never meet most of these people, I don't know what devices these people uses. I just trust they can find our SSID : 'myhotel', connect to it and that the captive portal login page shows up.
They can enter a 3 digit room number - and a password that is shown in the room.
And that's all. People can use our captive portal.

It doesn't "break" my company network, which lives on other local networks like 192.168.1.1/24 and 192.168.3.1/24. Why should it ?

My portal configuration is a little bit above 'bare minimum' as I choose to use a https portal login page. This needs a domain name for local usage, and the pfSense acme package.
It's a bit silly to protect the connection to portal web server this way, as no sensitive information is to be hidded : the password and room number are nearly public info. IPs used are RFC1918.
The thing is : most browsers want a trusted (= trusted web server certificate) https connection these days. They will yell when a http is used.

I'm using also the FreeRadius package and a mysql server 'scratch-pad' because "why not ?". For my usage, I can very well use the local pfSense user manager, making my pfSense system way easier to maintain.

I've mentioned the "classic Youtube captive portal video's from Netgate" above.
With the instructions from these videos, you should have a captive portal working.
If not .... well, the trouble shoot page should tell you that your DNS setup isn't what you want for a captive portal.
The pfSense default DNS setup - is mandatory - if you want the portal to work.

The captive portal trouble shoot page should be used to troubleshoot ;) If you find a something not clear, then that info should be used to post questions here.
Questions with details, of course.

Details are needed, so the cycle "you say it doesn't work" and "I say it does work" can be broken.
Otherwise I have to come over and redo your portal setup myself and call it a day. ( I'm joking of course).
Keep in mind : we all use the same software. Only our settings are different.

[DamienGombaultRecia]

Hi.

Some bugs related to captive portal with version 22.05 were reported on the pfSense Redmine issue tracker.
The most interesting one is : https://redmine.pfsense.org/issues/13488

Captive Portal: All users are given the same dummynet pipe pair
The reporter tells :

[...]
It also means that when a user is disconnected the pipe pair is deleted and all users are unable to pass traffic.
This can happen either via a manual disconnect or via a timeout.
This actually affects all users with or without bandwidth limiting set.
When there is no limit set all user are passed through an unlimited pipe but that still gets removed when one user is disconnected.

[Gertjan]

@damiengombaultrecia

Yep, that was this one ( see above ):

@gertjan said in Captive Portal is broken?:

See here : FreeRadius and quotas, doesn't work since 22.05 and if you use multiple portals (thus multiple interfaces) : Problem with multiple Interfaces since Version 22.05
For the last issue, there is a redmine present - and isn't even a patch proposed : just edit the file ( add a dot somewhere ;) )

[OpIT GmbH]

is there a Patch available?

[Gertjan]

@opit-gmbh

No official patches as of today.
If needed do the editing yourself.

[OpIT GmbH]

@gertjan

what i have to edit? i cant find it :=)

[Gertjan]

@opit-gmbh said in Captive Portal is broken?:

i cant find it :=)

Scroll upwards, see the second post if this thread for the 'bas switch!' patch.
See this one for another patch.

Also look post number 12.
You will have to edit the /etc/inc/captiveportal.inc to apply it.

If you do not use (FreeRadius) quotas, don't bother patching.

[OpIT GmbH]

@gertjan

and which one should i use? Iam not using any Radius Setup. Just Bandwidth limitations...

[Gertjan]

@opit-gmbh

Open this file : /usr/local/captiveportal/index.php and locate this line ( around line 251)

	$pipeno = captiveportal_get_next_dn_ruleno('auth', 2000, 64500, true);

change true for false :

	$pipeno = captiveportal_get_next_dn_ruleno('auth', 2000, 64500, false);

save.

Now, pipes are working per user as it should be.

Btw : this my quick and dirty hack. I'm using this 'false' for several weeks no : it works.

I see now Limiters and Schedulers per user under Diagnostics >Limiter Info.

[OpIT GmbH]

@gertjan

THX, it seams to work.

Limiters:
02002: 2.048 Mbit/s 0 ms burst 0
q133074 100 sl. 0 flows (1 buckets) sched 67538 weight 0 lmax 0 pri 0 droptail
sched 67538 type FIFO flags 0x0 16 buckets 0 active
02003: 2.048 Mbit/s 0 ms burst 0
q133075 100 sl. 0 flows (1 buckets) sched 67539 weight 0 lmax 0 pri 0 droptail
sched 67539 type FIFO flags 0x0 16 buckets 0 active
02000: 2.048 Mbit/s 0 ms burst 0
q133072 100 sl. 0 flows (1 buckets) sched 67536 weight 0 lmax 0 pri 0 droptail
sched 67536 type FIFO flags 0x0 16 buckets 0 active
02001: 2.048 Mbit/s 0 ms burst 0
q133073 100 sl. 0 flows (1 buckets) sched 67537 weight 0 lmax 0 pri 0 droptail
sched 67537 type FIFO flags 0x0 16 buckets 0 active

I see here 4 Limiters but i just have 2 Captive Portal Users, is this OK? Is this because one is for Upload and one for Download or?

[Gertjan]

@opit-gmbh
For each user, one down and one upload limiter.

[OpIT GmbH]

@gertjan

cool thx.

here is my Patch what iam using under System > Patches

diff --git a/opit/usr/local/captiveportal/index.php b/opit/usr/local/captiveportal/index.php
--- a/opit/usr/local/captiveportal/index.php
+++ b/opit/usr/local/captiveportal/index.php
@@ -249,7 +249,7 @@
$context = 'first';
}

• $pipeno = captiveportal_get_next_dn_ruleno('auth', 2000, 64500, true);

• $pipeno = captiveportal_get_next_dn_ruleno('auth', 2000, 64500, false);
  /* if the pool is empty, return appropriate message and exit */
  if (is_null($pipeno)) {
  $replymsg = gettext("System reached maximum login capacity");

[hsrtreml]

@gertjan said in Captive Portal is broken?:

@denx

Hi, I'm using 22.05 on a SG 4100. In the past, I was using 2.6.0 on a home made pfSense device.

If my captive portal wasn't working, I would lose my job - sort-of, as I use the portal for a hotel.

Last time I looked, it worked.

I'm using a dedicated interface for my portal, and as LAN has already 192.168.1.1/24, my portal uses 192.168.2.1/24 - DHCP pool 192.168.2.5 -> 254.

I use the Resolver, nearly default settings.

The pfSense patches packages has one patch for the portal :
This is the patch ID to be used : https://github.com/pfsense/pfsense/commit/add6447b9dc801144141bb24f8c264e03a0e7cae.patch

🔒 Log in to view

after install patch - same error
dummynet: bad switch 21!
dummynet: bad switch 21!