CCR1036-8G-2S+ all pppoe disconect every 3-4 days

I have a CCR1036-8G-2S+ 500pppoe clients , no bridge configuration on it , but every 3-4 days all pppoe get disconnected . Traffic 800-900mbps

All at the same time ?
So your CCR acts as a PPPoE server ?

Yes all at the same time .

and then reconnect in 3-5 min . and crazy think is dosn’t do always on peak hours , yesterday was done on 07:30AM bot most of the time 19:00-23:00 pm

Anything in the Log ?
Did you try enabling debug && PPPoE in the logs to see if you get any information ?
Or maybe it would be best if you monitored your router with an SNMP Software and collect valuable info such as CPU, Memory utilization etc. so that you can see the Router’s status at the time the problem appears…

i have enable cpu graphing but nothing show there …

i have add log with 10 000 lines and no error show there just x pppoe disconnected for all pppoe clients than connected

How to debug the pppoe please ?

My suggestion is an external SNMP monitor…

It is better to use a remote syslog server to collect your logs…

system logging add topics=pppoe,debug action=remote

pppoe get disconnected right now again …




nothing show on log , just hang up /disconnected and then connected again …

make an export of your configuration, then remove any private info on it and share it with us to try an assessment

CODE

i see a possible Layer 2 misconfiguration case like those documented by MikroTik:

follow recommended configuration guidelines on following articles

https://help.mikrotik.com/docs/display/ROS/Layer2+misconfiguration#Layer2misconfiguration-VLANinabridgewithaphysicalinterface

https://help.mikrotik.com/docs/display/ROS/Layer2+misconfiguration#Layer2misconfiguration-BridgedVLAN

Since v6.41 this kind of bridge configuration is no longer valid

First of all Disable STP, RSTP on bridge interface, that can help a little with some flapping

is very important to reconfigure your VLAN and bridge using bridge VLAN filtering

1st of all export should be done without password. Do we need to say thank you for giving us your L2TP Preshared Key…

thank you i test it , lets hope it will work .

Thank you , they are not real i randomly edited them before posting :slight_smile:

ok, please confirm if that resolves the issue or not

they just disconnected right now , i have submit ticket to mikrotik support 6 days now no response from them , i am thinking its hardware problem .. :frowning:

i suppose you have double checked access network and any other network element, passive or active between the CCR and PPPoE Client

i have many scenarios with CCR and many PPPoE Clientes working ok in a variety of scenarios:
Bras PPPoE + connection tracking off + fast-path mode
Bras PPPoE + connection tracking on + NAT + Fast-track mode
Bras PPPoE + connection tracking on + simple queues
Bras PPPoE + connection tracking on + simple queues + QoS per customer
Bras PPPoE + connection tracking on + simple queues + QoS per customer + NAT

I find very difficult faulty router having this kind of failure without leaving a trace of some problem on it or other equipment or totally crashing

at the same time not leaving a trace on other equipment

i think you need more monitoring on the rest of the network to prove is a CCR problem

maybe you are chasing the simptom, without knowing or finding the root cause of the problem, this is very common in PPPoE scenarios

i have two other ccr one with 700 clients and other one with 350 . they are working fine .

i had a ccr1009 and changed with this one 2 month ago .. 1009 was working completely fine . 1036 from the time we power on till today same thing . nothin else is changed on infrastructure only ccr . migration was done with export import command …if you have time i can give you access …

i have installed ccr 1009 day 2 everything seems to work fine (cpu 60%)

Mikrotik support claim that i have package losses but that’s not true because i have good ping with all devices on my network (0-1ms)

I am sure now that is a hardware problem , they just don’t want to admit.

Did you try to netinstall the device and configure again ?

No , do you think that will do any change ?

I don’t think so coz i have downgraded to max and updated to 7.1.2 still the same think …