部分网络命令

发表于 分类于
介绍 ifconfig 查看网卡参数，ss 与 netstat 查看套接字，以及 netstat -s 解读协议栈收发统计信息

设置 ip 地址

1
2
ifconfig eth0 192.168.1.10
ifconfig eth0 192.168.1.10 netmask 255.255.255.0

参数解释

比如 查看 网卡 eth0

1
2
3
4
5
6
7
8
9
10
[root@imwl-175 ~]# ifconfig eth0
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet <PUBLIC_IP>  netmask 255.255.254.0  broadcast <PUBLIC_IP>
        inet6 fe80::xxxx:xxxx:xxxx:xxxx  prefixlen 64  scopeid 0x20<link>
        inet6 <PUBLIC_IP>  prefixlen 64  scopeid 0x0<global>
        ether xx:xx:xx:xx:xx:xx  txqueuelen 1000  (Ethernet)
        RX packets 3193275  bytes 2157200207 (2.0 GiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 3225664  bytes 2355880326 (2.1 GiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
  1. RUNNING : 表示网络是连通的
  2. MTU : 默认为 1500
  3. TX : 发送
  4. RX : 接受
  5. errors : 错误的数据包数，eg: 校验错误，帧同步错误
  6. dropped : 丢弃的数据包数，数据包已经到了 ring buffer, 但因为内存不足等原因丢包
  7. overruns : 超限数据包， 网络 IO 过快，导致 ring buffer 中的数据包来不及处理（队列满）而导致的丢包
  8. carrier : 一般是双工不匹配，物理电缆出现问题
  9. collisions : 碰撞数据包数

ss 与 netstat 查看套接字信息

-l : 监听套接字
-n : 显示数字地址和端口
-p : 显示 进程信息
-t : 显示 tcp 套接字

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
[root@imwl-175 ~]# netstat -tlunp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp        0      0 0.0.0.0:111             0.0.0.0:*               LISTEN      537/rpcbind
tcp        0      0 0.0.0.0:80              0.0.0.0:*               LISTEN      919/nginx: master p
tcp        0      0 0.0.0.0:22              0.0.0.0:*               LISTEN      909/sshd
tcp        0      0 127.0.0.1:25            0.0.0.0:*               LISTEN      1038/master
tcp        0      0 0.0.0.0:443             0.0.0.0:*               LISTEN      906/nginx
tcp6       0      0 :::111                  :::*                    LISTEN      537/rpcbind
tcp6       0      0 :::8080                 :::*                    LISTEN      4307/docker-proxy-c
tcp6       0      0 :::22                   :::*                    LISTEN      909/sshd
tcp6       0      0 ::1:25                  :::*                    LISTEN      1038/master
tcp6       0      0 :::2377                 :::*                    LISTEN      3586/dockerd-curren
tcp6       0      0 :::7946                 :::*                    LISTEN      3586/dockerd-curren
udp        0      0 0.0.0.0:68              0.0.0.0:*                           846/dhclient
udp        0      0 0.0.0.0:111             0.0.0.0:*                           537/rpcbind
udp        0      0 127.0.0.1:323           0.0.0.0:*                           590/chronyd
udp        0      0 0.0.0.0:4789            0.0.0.0:*                           -
udp        0      0 0.0.0.0:700             0.0.0.0:*                           537/rpcbind
udp6       0      0 :::7946                 :::*                                3586/dockerd-curren
udp6       0      0 :::111                  :::*                                537/rpcbind
udp6       0      0 ::1:323                 :::*                                590/chronyd
udp6       0      0 :::700                  :::*                                537/rpcbind

Recv-Q : 接收队列
Send-Q : 发送队列

协议栈信息

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
[root@imwl-175 ~]# netstat -s
Ip:
    3215047 total packets received
    6537 forwarded
    1 with unknown protocol
    0 incoming packets discarded
    3208496 incoming packets delivered
    3295252 requests sent out
    7 outgoing packets dropped
Icmp:
    1434 ICMP messages received
    48 input ICMP message failed.
    ICMP input histogram:
        destination unreachable: 472
        timeout in transit: 11
        echo requests: 947
        echo replies: 4
    3204 ICMP messages sent
    0 ICMP messages failed
    ICMP output histogram:
        destination unreachable: 2257
        echo replies: 947
IcmpMsg:
        InType0: 4
        InType3: 472
        InType8: 947
        InType11: 11
        OutType0: 947
        OutType3: 2257
Tcp:
    28888 active connections openings
    102936 passive connection openings
    3978 failed connection attempts
    2579 connection resets received
    7 connections established
    3006869 segments received
    3442563 segments send out
    287671 segments retransmited
    42693 bad segments received.
    54447 resets sent
    InCsumErrors: 42679
Udp:
    342293 packets received
    2261 packets to unknown port received.
    0 packet receive errors
    340842 packets sent
    0 receive buffer errors
    0 send buffer errors
UdpLite:
TcpExt:
    859 resets received for embryonic SYN_RECV sockets
    79 packets pruned from receive queue because of socket buffer overrun
    1 ICMP packets dropped because they were out-of-window
    23297 TCP sockets finished time wait in fast timer
    168067 delayed acks sent
    12 delayed acks further delayed because of locked socket
    Quick ack mode was activated 23425 times
    2 SYNs to LISTEN sockets dropped
    506703 packet headers predicted
    1088634 acknowledgments not containing data payload received
    510495 predicted acknowledgments
    23673 times recovered from packet loss by selective acknowledgements
    Detected reordering 13094 times using SACK
    Detected reordering 29 times using reno fast retransmit
    2 congestion windows fully recovered without slow start
    464 congestion windows recovered without slow start by DSACK
    578 congestion windows recovered without slow start after partial ack
    TCPLostRetransmit: 27128
    242 timeouts after SACK recovery
    250 timeouts in loss state
    155981 fast retransmits
    2415 retransmits in slow start
    21238 other TCP timeouts
    TCPLossProbes: 31134
    TCPLossProbeRecovery: 4774
    2579 SACK retransmits failed
    23720 DSACKs sent for old packets
    49 DSACKs sent for out of order packets
    9217 DSACKs received
    339 DSACKs for out of order packets received
    1380 connections reset due to unexpected data
    46 connections reset due to early user close
    13845 connections aborted due to timeout
    TCPDSACKIgnoredOld: 55
    TCPDSACKIgnoredNoUndo: 5132
    TCPSpuriousRTOs: 92
    TCPSackShifted: 137476
    TCPSackMerged: 343569
    TCPSackShiftFallback: 165507
    TCPRcvCoalesce: 244304
    TCPOFOQueue: 18977
    TCPOFOMerge: 49
    TCPChallengeACK: 37
    TCPSYNChallenge: 14
    TCPFastOpenActive: 1681
    TCPFastOpenActiveFail: 1445
    TCPFastOpenCookieReqd: 848
    TCPFastOpenBlackhole: 5
    TCPSpuriousRtxHostQueues: 28
    TCPAutoCorking: 20068
    TCPFromZeroWindowAdv: 853
    TCPToZeroWindowAdv: 855
    TCPWantZeroWindowAdv: 657
    TCPSynRetrans: 8587
    TCPOrigDataSent: 2429980
    TCPACKSkippedSynRecv: 16
    TCPACKSkippedSeq: 642
    TCPACKSkippedChallenge: 2
    TCPWinProbe: 34
    TCPKeepAlive: 626
    TCPWqueueTooBig: 17
IpExt:
    InNoRoutes: 13
    InOctets: 1219007134
    OutOctets: 2346546494
    InNoECTPkts: 3217097
    InECT1Pkts: 1
    InECT0Pkts: 450
    InCEPkts: 1801

这段输出怎么读

不用逐行看，挑几个能算出结论的比值就够了。netstat -s 给的是开机以来的累计值，所以要么两次采样做差，要么直接用 nstat（默认只报增量）——单看一次快照，看到的是总账，不是现在的状况。

这台机器的数据恰好很有戏：

① 重传率 = segments retransmited / segments send out = 287671 / 3442563 ≈ 8.4%

正常网络应该在 0.1% 这个量级，8.4% 说明链路或者对端有实质问题。这是判断”网络到底有没有毛病”最该先看的一个数。

② 坏包里几乎全是校验和错误 = InCsumErrors / bad segments received = 42679 / 42693 ≈ 99.97%

坏包本身有很多来源（截断、序号异常、选项非法），但这里 42693 个坏包中有 42679 个是校验和错——方向就很明确了，指向物理链路或网卡（网线、光模块、交换机端口、offload 出问题），而不是软件逻辑。配合 ethtool -S <iface> 里的 rx_crc_errors 一起看基本能定死。

③ TCP Fast Open 失败率 = TCPFastOpenActiveFail / TCPFastOpenActive = 1445 / 1681 ≈ 86%

典型症状是路径上有中间设备（防火墙、负载均衡、某些运营商设备）把 TFO 选项吃掉了。这种情况下开着 TFO 只会白多一轮握手延迟——失败后要退回普通三次握手，比不开还慢。要么确认路径支持，要么关掉。

方法论上，这三个比值分别对应三类判断：分子分母都在同一层才有意义（别拿 IP 层的丢包去除 TCP 的发送数）；比值比绝对值可靠（绝对值受运行时长影响）；能定位到层的比值最值钱（②直接把问题从软件推到了硬件）。