search for: li3

Hi, I have three linux machines, and I want to let one of them forward packets betwen the other two. The forwarding node has two ethernet cards, connecting the two two machines respectively. However, when I ping between the two end points, the forwarding node can receive the ping requests at its eth0, but it never forwards them to its eth1. So is the reverse direction. The forwarding node is

OK, I spammed the mailing list recently, but I will be fired if I can''t solve the problem today. (just kidding, but I did waste lots of time on it :( ) The common configuration for teql is for two computers connected directly with two links. My topology is a little different: one link is connected directly, but the other is connected through a gateway. My problem is teql can''t

Hi all, I want to implement a "weighted" teql that can send packets to interfaces based on their "weights". To do this, I want to create multiple virtual network interfaces, and add them to teql. Since teql will send packets to each interface in a round-robin way, a weighted teql is achieved (my guess). I already have two physical network interfaces, eth0 and eth1. I

Hi, I want to study the multi-path TCP behavior (single TCP connection over multi-paths), and I am setting up the test bed. I have two machines, each with two network cards. I have directly connected two eth0, and I am trying to figure out how connect two eth1 to control the properties of this connection, so that I can study the TCP on multiple different paths (different in terms of latency,

Thanks in advance! Summary: when I load netem and teql together, teql doesn''t work correctly. (If I load teql only, everything is fine) I loaded both netem and teql. Netem is associated with eth0, and teql is associated with both eth0 and eth1. But traffic only goes out of eth1. Attached are the commands that I used to configure teql and netem (on machine 1), and commands to

I am measuring the performance of one TCP connection over two symmetric paths. Packets are sent to two paths alternatively. I found that when the latency of each path are within 1ms, the overall TCP throughput is the *sum* of the throughput of the two paths. However, when the latency of the two paths increases to 5ms, the overal TCP throughput drops to the throughput of a *single* path. Has anyone

Yesterday I posted a question, but I guess too much detail is provided that no one would bother to read, so I rephrase the question and hopefully some one may be willing to read. Sorry to spam. I am using TEQL on two computers, each with two network interfaces. The two eth0 are connected directly, and the two eth1 are connected via a gateway (a linux machine). The problem is that when I send

TEQL and Subnet problem I have a network topology shown below, and I am trying to use TEQL. P1, P2 and P3 are three linux machines, and each has two network cards. P2 and P3 are connected in two ways: (1) directed connection between P2.eth0 and P3.eth0; (2) connection via P1. P1 acts as the gateway. TEQL is installed on P2 and P3 to load-balance packet transmission by round-robin sending packets

TEQL and Subnet problem I have a network topology shown below, and I am trying to use TEQL. My problem is: When I ping to P3.teql0 from P2 ("[P2]# ping 16.119.144.66"), the traffic can never go from P2.eth1, and all traffic only goes to P1.eth0. What P2.eth1 (16.119.144.33) did is broadcasting an ARP asking for the MAC address of 16.119.144.66, although I have specified the route to