Problems running PPPoE Server in v6.1

Hello all,

I’ve sent a lot of messages about the PPPoE problem when Mikrotik RoS is running as a NAS. The problem was reported in this post:

http://forum.mikrotik.com/t/problems-on-pppoe-over-different-beta-stable-versions/65617/1

After the stable connection we made and update over our NAS servers and we have 3 of them:

RB1100AH - CSC-auth-06
RB1100AHx2 - CSC-auth-08
RB1100AHx2 - CSC-auth-09

In all of those we got some PPPoE estabilished sessions but not running and, for that, our customer can’t have a connection because our OSPF, which is ment to share the IP over the network, do not share the IPs put on this “not running” connections.

To be better understod I am attaching a screen:

Can anyone help me with that? I am returning all the NASes to 6.0rc14 to see what happens.

I had same problem on 6.1 . here is official reply from mikrotik guys .

Hello,

Routing is crashing. We will fix this in future versions, but currently try not to
run OSPF on ppp interfaces, it might help.

Regards,
Maris

try this and report back http://mikrotik.com/download/share/routeros-x86-6.2rc1.npk

Thank you Omid. I will try this 6.2rc1.

Hi there guys,
i’m experiencing the same problem on 3 CC36 with RoS 6.2 installed.
Any news about this issue?

Thank you all in advance! :smiley:

Hello Max,

Can you enlight us with more information? How many PPPoE Sessions you can achieve before “Running No” became? Is there an ammount of traffic that you see it more closely?

Hi Nataniel,
thanks for your answer.
I have about 600 PPPoE sessions for each CC36.

After rebooting one of them i saw i got “not running” sessions after about 300 active sessions.

I tried to manually delete those sessions but in a matter of seconds they came back (not necessarily with the same name). I reached a maximum of 16 not running sessions at the same time.
I was forced to write down a script to automatically delete them but i don’t think it’s a great solution.

Still the same problem on our 3 x CCR1036 (OSPF + PPPOE over radius only) RouterOS 6.2, fw:3.09.

I already reported MT supports about this problem again and they only said: “Thank you very much for the report. We will research your issue.”

Next week or few days later we try to change our OSPF network type from ptp to ptmp, it is only intuition, i don’t know how it may help us.

Hi there…any news?
Hope at least this will be fixed in 6.3! :slight_smile:

Ehi guys…just tried to install 6.3 and it seems to work fine! Not-running pppoe clients don’t even show up and the authentication process of the pppoe sessions is faster.
Finally! :smiley:

I think we must wait one or two weeks of uptime before rejoice.
Invalid IP’s don’t appear immediately after reboot.