查看网速

ethtool eth0 | grep -i Speed
        Speed: 1000Mb/s
        
apt install sysstat
sar -n DEV  1

11:48:54        IFACE   rxpck/s   txpck/s    rxkB/s    txkB/s   rxcmp/s   txcmp/s  rxmcst/s
11:48:54         eth0     17.07     12.20      1.23      1.86      0.00      0.00      0.00
11:48:54           lo      0.00      0.00      0.00      0.00      0.00      0.00      0.00

Average:        IFACE   rxpck/s   txpck/s    rxkB/s    txkB/s   rxcmp/s   txcmp/s  rxmcst/s
Average:         eth0     20.91     22.80      3.67     18.55      0.00      0.00      0.00
Average:           lo      1.49      1.49      0.39      0.39      0.00      0.00      0.00

Linux Performance Observability: sar

ss

Another utility to investigate sockets

基本用法

ss -t -a
      Display all TCP sockets.

ss -u -a
      Display all UDP sockets.

ss -o state established '( dport = :ssh or sport = :ssh )'
      Display all established ssh connections.

ss -x src /tmp/.X11-unix/*
      Find all local processes connected to X server.

ss -o state fin-wait-1 '( sport = :http or sport = :https )' dst 193.233.7/24
      List all the tcp sockets in state FIN-WAIT-1 for our apache to network 193.233.7/24 and look at their timers.

ss -i 详解

Netid  State      Recv-Q Send-Q   Local Address:Port       Peer Address:Port
ESTAB       0      0 ::ffff:127.0.0.1:2181 ::ffff:127.0.0.1:52150
cubic wscale:6,7 rto:204 rtt:7.5/3 ato:40 mss:21888 cwnd:10 ssthresh:7 send 233.5Mbps rcv_rtt:22200.7 rcv_space:43690
  • Recv-Q，The count of bytes not copied by the user program connected to this socket.
  • Send-Q，The count of bytes not acknowledged by the remote host.
  • cubic，TCP 拥塞控制使用的算法，通过 /proc/sys/net/ipv4/tcp_congestion_control 设置
  • wscale，rwnd 的 scale 值，逗号前面的 6 对远端提供的 rwnd 起作用，逗号后面的 7 对我们提供给远端的 rwnd 起作用。/proc/sys/net/ipv4/tcp_window_scaling 控制是否开启 scale
  • rto，Retransmittion Timeout，204ms 内没收到 ack，理论上最小为1s，但 Linux 实现中默认一般小于1s，见参考5。 netstat -s | grep retrans 可以看到重传数
  • rtt，Round Trip Time (is the delays between sent packets and received ACK)
  • cwnd，current congestion window size
  • ssthresh，a system-configured congestion threshold。在 tcp 慢启动阶段，cwnd <= ssthresh 。在 /sys/module/tcp_cubic/parameters/initial_ssthresh 中设置，默认为 0.
  • rcv_rtt，The time to receive one full window
  • rcv_space，twice the most recently measured window? TCP's internal auto-tuning to grow socket buffers based on how much data the kernel estimates the sender can send

参考：

  1. http://www.misterx.org/wiki/index.php/Ss
  2. https://github.com/jiacai2050/ideas/issues/41#issuecomment-362791333
  3. What is rcv_space in the 'ss --info' output, and why it's value is larger than net.core.rmem_max
  4. https://blog.janestreet.com/inspecting-internal-tcp-state-on-linux/
  5. Performance Best Practice: Monitor TCP Retransmits
  6. http://www.lognormal.com/blog/2012/09/27/linux-tcpip-tuning/

ss 源码

  • https://git.kernel.org/pub/scm/network/iproute2/iproute2.git/tree/misc/ss.c