we are reporting in SUP-136050 that after a while of regular functioning the serer pppoe crashes and let a lot of queue and dhcpv6 server on unknow interfaces.
The problem could be due to a duplicated interface id but Mikrotik seems unable to find it.
Almost all the time the server crashes it generate a supout and I update the ticket about it… tons of supout not audited by MT.
Please MT fix the issue… it is only three years old!!!
I take it that you have not found a reliable way to reproduce the problem?
What if you were to set up a ROS 7 PPPoE server in a lab, make hundreds of test connections to it (you could theoretically do this from a single device), and then artificially cause a lot of “churn”, which would force the active sessions to constantly reconnect? Does the problem trigger quickly if you do something like that?
we are not able to reproduce it into the lab.
We get it on ccr2216 and x86 platform too.
I definitely get that it’s frustrating, but if you can’t reproduce it, then MT probably also can’t reproduce it.
What have you tried to do to reproduce? If the theory is that interface ID for a dynamic PPPoE client interface eventually picks a duplicate and collides, it seems like you can increase the odds of this happening with many connections and disconnections happening per second simultaneously, for example.
after several tents of supout…