in my CCR1036 I still have the same problem when it drops more than 500 pppes cpu of a core is at 100%, before it was not like that it was after upgrading to version 6.47 long term
Do you do NAT on the same box and do you have a “masquerade” rule for that? If so, replace it by a suitable “src-nat” rule.
I took all the “masquerade” and put netmap but the problem still remains, when they disconnect more than 500 pppoes it seems that the cpu processing increases, I forgive access to CCR1036 and when these 500 pppoe are connecting again at the same time the cpu increases and the CCR1036 is unavailable until all pppoes go up, I removed all the “masquerade” rules, but the problem remains.
[*]Hello all
I have same issue when i kick 10 pppoe users on my 1036 8G2s+ it drops 200-250 pppoe users out of 450 active. This should be the mikrotik bug i do everything that ı know and ı read about but no solution. I have src nat and netmap on it
same router 1036 8G 2+ having this issue. Anybody found solution?
I still had the problem when disconnecting more than 200 pppoe, forgive access via winbox to CCR1036, I’ve checked the NAT does not have any masquerade rule, I have 2 CCR1036 and it’s the same someone already went through this, whenever more than 200 connections drop pppoe I get no access and CCR1036 traffic drops
Hi
How can I send winbox’s file from Mikrotik to windows without download it ?
Is it possible?
I have created a new post with this because it is marked as SOLVED, and it isn’t like that.
The problem is that setting up new PPPoE connections or tearing down existing PPPoE connections requires much more CPU attention than merely forwarding traffic.
So when you have 2000 connections the CCR can happily handle them, but when 200 are closed or newly opened at once the CPU is completely overwhelmed.
At some point so badly that it also cannot maintain the existing 1800 connections and an avalanche occurs in which all connections are broken and there is no CPU available to handle new connections either, and you have to hope that the customers slowly make their connection again instead of all at once.
This scenario is not unique, it can happen everywhere in IT systems, e.g. also with websites that get too much traffic.
@pe1chi. thanks so much for your reply, but I just moved my answer to a new Post because this is marked as Solved.
Here is the URL: http://forum.mikrotik.com/t/pppoe-connection-was-already-active-closing-previous-one/153566/1
The same thing happens on the mikrotik 2216. There is no other active connection. this is a bug mikrotik V 7.14.3
/ppp active remove [find name="$user"]
:log warning "PPPoE User $user disconnected and active session removed."
try to add this script into on-down in profile script
/ppp active remove [find name="$user"]
:log warning "PPPoE User $user disconnected and active session removed."
I think I’ve already solved the problem. It seems that the issue was caused by two devices having the same MAC address, which was causing them to conflict with each other.
What you need to do is go to Active Connections and get the Caller ID of the client. Then, go to Secrets and add the client’s Caller ID there. This will make the Caller ID static for that client, preventing conflicts and stopping the connections from competing with each other.
Basically, all PPPoE accounts in the Secrets should have a static MAC address so that the router won’t have to search for the MAC address again if they disconnect, preventing other clients from being affected.