I have the same problem!! The only difference is when I have no intermediate router works!! The costumer has a complete access. But when i try to implement the vpls with a costumer using either intermediate router or bonding the routing for outside network doesn’t work, only the internal routing.......

I have the same problem!! The only difference is when I have no intermediate router works!! The costumer has a complete outside access. But when i try to implement the vpls with a costumer using either intermediate router or bonding the routing for outside network doesn’t work, only the internal rou...

We migrate our network from OSPF/EOIP to a MPLS/VPLS solution seeking for a better QOS management. The issue it’s not on ingress routers. The proper set on EXP field it’s done at the bridge filters. We set the exp 7 for voice and 0 for the other traffic. That set is forward to the core router but I ...

if you change any conf at this router (NKT 1000) Something make the router reboot with "System Rebbotted because kernel failure"

you are nating your clients ? If yes you have a problem, cant use my filters because every connection gos to the router then nated.... Try using two diferent routers or interfaces and routing betwen then. Using the rules on the external e nating the internal...Its a idea...

its internal BGP or external ?

what chain forward or input ?

to(input) the router with more then 10 conections per second its a DOS for sure!!!! P2P and users connections go Throug the router (forward)!!!! syn its another kind of atack.... you can deal with that to. Limit for no more then 25 TCP syn per second (forward and input) But what you whant prevent a ...

the correct rules are add chain=input action=tarpit protocol=tcp src-address-list=Add_Ddos \ comment="TARPIT new connections from the dos source ip" disabled=no add chain=input action=add-src-to-address-list protocol=tcp \ connection-limit=10,32 src-address-list=!IPsListaBranca \ address-l...

Your idea its ok but the interface to attach its the global-total. Because you want to prioritize the traffic trough the router and always put limits because if you don’t do the mkt thinks “ok no limits nothing to do...”

Dagmik

try

2 chain=bgp_crowley_out prefix=(your network) invert-match=yes action=discard

Jean

the same with ver 2.9.40 in simple queues and VRRP address, and inerface address.

Yah, yah now bgp best path algorithm works but with 3 peers and the return off unused routes to table, MK needs to handle 600.000 lines... And reset all three peers after bout 3 hours. Already open ticket but dudes please take BGP serious... We try to convince people that your solution is reliably b...

Today I test the 2.9.41 apparently the refresh bug is solved but the BGP session can’t stay steady... After 2 hours reset...the route table process is absolutely slow. About 1 hour to process 2 peers full table. I think that’s no read yet... Last but not least the table process take all the processo...

We have the same refresh problem described in the previous post. Already send a email to support with the support files but any news about? If the mikrotik support wants ,its possible to provide access to our MKT core backup router to make some tests. Only to describe for the other topic readers: We...

AS a large ISP we use a MKT as core with BGP and deploy all clients with the same DELL switch. With one use a VLAN so we have 68 VLANS. Try 2 steps: 1 You update the DELL OS switch ? 2 Thats was our problem, are you check the NIC compatibility list for VLANS at mkt site ? Our model are the D-LINK gi...