Page 1 of 1

PPPoE - invalid queues and active connections problem

Posted: Thu Oct 25, 2012 9:21 pm
by fabricioviana
Hi! I have this strange problem.

I have a RB493G with some PPPoE clients with FreeRadius.

I was looking in PPP > Active Connections tab when I saw that one user had 3 connections with same Session ID, IP and Uptime!

Then I changed to PPP > Interface and there was only one interface with this pppoe-user. I disconected this user.

When I came back to PPP > Active Connections and 1 connection disappeared but the other 2 became totaly messed! The uptime was 5 minutes jumped to more than one day and the MAC ADDRESS became totally crazy! Please look at problem1.jpg attached.

Then I change to Queues and they were all messed too, all invalid! Please, look at problem2.jpg attached.

Then I change back to Active Connections and there wan no more connections. Same in Interface.

Because these invalid queues MK does not allow user do relogin.

I solve this deleting all queues. Then everybody reconnect!

Before RB 493G I had a RB 450G and the same problems occurred...

Some times this problem occurs many times at same day. Some times take days to occour!

Has anyone experienced this problem?

Best Regards
Fabricio

Re: PPPoE - invalid queues and active connections problem

Posted: Fri Oct 26, 2012 7:25 am
by nmthaker
Hello

Between your RADIUS and NAS there is always connection connected but if some time it drops and the same time session time out not sent to the NAS the NAS wont connect the user or RADIUS wont connect the user due to the same you can able to see that user online on your web portal

If you made any changes which affect RADIUS but you didnt not restart RADIUS then also this problem occurred. This is nothing serious. you can ignore it or once you make some major changes in system do RESTART RADIUS server or even you can manually remove all clients from NAS

Regards

Nishit

Re: PPPoE - invalid queues and active connections problem

Posted: Fri Oct 26, 2012 10:01 pm
by fabricioviana
Thanks for your reply!

I´ll try, in that order:
- increase keepalive timeout (today is 10 secs).
- increase keepalive timeout and mark one session per host.
- decrease keepalive timeout to zero and mark one session per host.

If one of these solves my problems I´ll post here.

I kept keepalive timeout in 10 seconds and checked the one session per host option but the problem occured again.

Any other ideias?

Regards
Fabricio