Hello,
I’ve installed a Mikrotik instance on a server a few weeks ago with the sole purpose of having an easy to use traffic shaper. The setup was rather simple: I’ve plugged in ether1 in a Cisco port set in access, and plugged ether2 in a Linksys port. The Linksys port is where I plugged in a few machines. I’ve set up the transparent bridge, created the queues and everything is working just fine.
Now I want to put IPs from another VLAN behind the traffic shaper. After all it’s an Intel server with a Xeon CPU with 2 gigabit NICs so handing a traffic for three OpenVZ and Xen nodes is a simple waste of processing power. My plan is to have one traffic shaper in each rack that will shape the traffic for a couple of servers each.
I’ve setup an IP address from another VLAN (not the one initially set on the Cisco port in access mode), I’ve switched the Cisco port from Access to Trunk, set the initial VLAN as native and allowed two more (the test IP belongs to one of the other two VLANs).
My question is: where do I define these two additional VLANs? On ether1? On ether2? On bridge 1? On the Linksys port that the Mikrotik server’s ether2 is plugged into? On the Linksys port where the node holding the test VPS machine is plugged in? On more than one places? If yes which? Which one will be tagged and which one will be untagged?
I’m rather new with these things, I have a simple idea of how the packets move across the networks, where are they encapsulated and where de-encapsulated, however all my tests lead nowhere, no matter what I done, I simply can not ping the test IP.
I’d apreciate and idea from more experienced users.
Thanks.