[RST] was sent on the second subflow immediately after [SYN, ACK] was received---almost the same with the issue164

leidemon commented 4 years ago

hi, mptcp team: I use the mptcp to do some test,but i face one issue like issue164,can you give me some suggestions? use: ubuntu 16.04 mptcp 0.95 shadowsocks-libev-3.3.3 I use one PC for ss-server,config as follow: { "server":"0.0.0.0", "server_port":8388, "local_port":1080, "password":"123456", "timeout":600, "method":"aes-256-cfb", "verbose":2, "mode":"tcp_and_udp" } cmd: ss-server -c config.json --mptcp -u one for the ss-local(with privoxy for agent),configure as follow: { "server":"192.168.9.42", "server_port":8388, "local":"127.0.0.1", "local_port":1080, "password":"123456", "timeout":600, "method":"aes-256-cfb", "mode":"tcp_and_udp", "verbose":2 } cmd:ss-local -c config.json --mptcp -u and the privoxy conf: listen-address 127.0.0.1:8118 forward-socks5 / 127.0.0.1:1080 . ss-local have two interface: 192.168.9.43 192.168.9.199 when I test for mptcp aggregate Download,[RST] was sent on the second subflow immediately after [SYN, ACK] was received, would some one else can give me some suggestions?

leidemon commented 4 years ago

mptcp config: net.mptcp.mptcp_path_manager = fullmesh net.mptcp.mptcp_scheduler = redundant net.mptcp.mptcp_enabled = 1

leidemon commented 4 years ago

in server side,the MPCapableFallbackSYNACK is increasing for each connect;

cpaasch commented 4 years ago

Please take a look at my comment in https://github.com/multipath-tcp/mptcp/issues/164 - Can you gather individual pcaps on each interface? Just to confirm my theory of asymetric routing.

leidemon commented 4 years ago

请查看我在＃164中的评论-您可以在每个界面上收集单独的pcap吗？只是为了确认我的非对称路由理论。 hi, cpaasch: I comment in issue164

leidemon commented 4 years ago

it can be closed @cpaasch

multipath-tcp / mptcp

[RST] was sent on the second subflow immediately after [SYN, ACK] was received---almost the same with the issue164 #375