Abnormal IP address in captive portal

[jimp]

Is this on 1.2.3 or 2.0?

[kloon]

it's on 1.2.3

We are not moving to a 2.0 beta version as this server is in production.

[kloon]

Hi all, the abnormal IP address appear again in our captive portal, do anyone have idea on it?

IP address    MAC address    Username    Session start   
Y2Fyb2x5bjk3OTE=                             01/01/1970 07:30:00

[cmb]

What are the contents of /var/db/captiveportal.db?

[kloon]

Would you mind to guide how to check the particular file?

[cmb]

Go to Diag>Command and run:

cat /var/db/captiveportal.db

[kloon]

sorry for the late reply, the problem does not appear again until today.

this is the content of /var/db/captiveportal.db

,cd4e0e5cac5b2548,Y2Fyb2x5bjk3OTE=,,,

[Gertjan]

'Normal' content is
$ cat /var/db/captiveportal.db
1281899223,10004,192.168.2.160,00:26:08:yy:xx:ii,105,405ce45ef655491c,,,,
1281899536,10005,192.168.2.173,d8:30:62:yy:xx:ii,105,e44437ff909ce274,,,,
1281902318,10006,192.168.2.150,90:4c:e5:ii:yy:xx,205,2742c8beaaa25b21,,,,
1281941548,10003,192.168.2.165,f8:1e:df:ii:yy:xx,114,a55ab8690b33af7a,,,,

(just an example of my captiveportal.db file right now)

Looks like the time stamp, IP, index and USER ID are missing.

Strange.

[kloon]

I have normal content in captive portal too, but sometimes the abnormal address appear in the portal randomly.

$ cat /var/db/captiveportal.db
,23c94761e47d958f,c2lsdmVyd2F0ZXI4,,,
,5b856ac5e01c8a86,cmhnanBhcmk=,,,
1282000190,10000,192.168.0.1,1c:4b:d6:11:22:33,abc123,4a8c88399ded7095,cmFmZjg5MDQyOQ==,,,
1282000315,10001,192.168.0.12,90:84:0d:44:55:66,def456,26fa9255b49fcc6a,eXBwaHBpZWU=,,,

[Gertjan]

Some how, the database $cpdb (list with all the connected clients) fails to write correctly to your disk.
Destroyed in memory ?

On what are you running pfSense ?

I would re-install from scratch.

[eri--]

How many concurrent users make this bug show??

[kloon]

It happen in a random number of users. Sometimes happen in 40+ users, sometimes it happen in 300+ users. But I do not sure when it happen since I refresh the Portal page once a while.

[kloon]

appreciate that anyone can help into this matter.

Thank you very much.

[cmb]

Not a common issue, we support loads of CP environments and never seen that. The back end handling of CP is much improved in 2.0, and it's throughly tested and used in numerous serious production environments (ISPs and others). You may want to try it and see if that changes anything. Somehow it's missing some data on the back end in rare circumstances, and that may well be fixed.

[kloon]

Thank you for your reply. We will try it out once stable version of 2.0 release.