tcpdump 中文man页面

名称 (NAME)

tcpdump - 转储网络上的数据流  

总览 (SYNOPSIS)

tcpdump [ -adeflnNOpqStvx ] [ -c count ] [ -F file ]

         [ -i interface ] [ -r file ] [ -s snaplen ]

         [ -T type ] [ -w file ] [ expression ]
 

描述 (DESCRIPTION)

Tcpdump 打印出 在某个 网络界面 上, 匹配 布尔表达式 expression 的 报头.

对于 SunOS 的 nit 或 bpf 界面: 要 运行 tcpdump , 你 必须 有 /dev/nit 或 /dev/bpf* 的 读访问 权限.

对于 Solaris 的 dlpi: 你 必须 有 网络仿真设备 (network pseudo device), 如 /dev/le 的 读访问 权限.

对于 HP-UX 的 dlpi: 你 必须 是 root, 或者 把它 安装成 root 的 设置uid 程序. 对于 IRIX 的 snoop: 你 必须 是 root, 或者 把它 安装成 root 的 设置uid 程序. 对于 Linux: 你 必须 是 root, 或者 把它 安装成 root 的 设置uid 程序.

对于 Ultrix 和 Digital UNIX: 一旦 超级用户 使用 pfconfig(8) 开放了 promiscuous 操作模式 (promiscuous-mode), 任何用户 都可以 运行 tcpdump.

对于 BSD: 你 必须 有 /dev/bpf* 的 读访问 权限.

选项 (OPTIONS)

-a

试着 把 网络和广播地址 转换成 名称.

-c

当 收到 count 报文 后 退出.

-d

把 编译好的 报文匹配模板 (packet-matching code) 翻译成 可读形式, 传往 标准输出, 然后退出.

-dd

把 报文匹配模板 (packet-matching code) 以 C 程序片断 的 形式 输出.

-ddd

把 报文匹配模板 (packet-matching code) 以 十进制数 形式 输出 (前面 加上 总数).

-e

每行 都 显示 链路层报头.

-f

用 数字形式 显示 '外部的' 互联网地址, 而不是 字符形式 (这个 选项 用来绕开 脑壳坏光的 SUN 黄页服务器 的 问题 --- 一般说来 它 翻译 外部网络数字地址 的时候 会 长期挂起).

-F

把 file 的内容 用作 过滤表达式. 忽略 命令行 上 的 表达式.

-i

监听 interface. 如果 不指定 接口, tcpdump 在 系统 的 接口 清单 中, 寻找 号码最小, 已经 配置好的 接口 (loopback 除外). 选中的时候 会 中断 连接.

-l

行缓冲 标准输出. 可用于 捕捉 数据 的 同时 查看 数据. 例如,
``tcpdump  -l  |  tee dat'' or ``tcpdump  -l   > dat  &  tail  -f  dat''.

-n

别把 地址 转换成 名字 (就是说, 主机地址, 端口号等)

-N

不显示 主机名字 中的 域名 部分. 例如, 如果 使用 这个 选项, tcpdump 只显示 ``nic'', 而不是 ``nic.ddn.mil''.

-O

禁止运行 报文匹配模板 的 优化器. 只有 当你 怀疑 优化器 有 bug 时 才有用.

-p

禁止 把 接口 置成 promiscuous 模式. 注意, 接口 有可能 因 其他原因而 处于 promiscuous 模式; 因此, '-p' 不能 作为 `ether host {local-hw-addr} 或 ether broadcast' 的 简写.

-q

快速输出. 显示 较少的 协议信息, 输出行 会 短一点点.

-r

从 file 中 读入 数据报 (文件 是用 -w 选项 创建的). 如果 file 是 ``-'', 就 读 标准输入.

-s

从每个 报文 中 截取 snaplen 字节的数据, 而不是 缺省的 68 (如果是 SunOS 的 NIT, 最小值是 96). 68 个字节 适用于 IP, ICMP, TCP 和 UDP, 但是 有可能 截掉 名字服务器 和 NFS 报文 的 协议 信息 (见下面). 输出时 如果指定 ``[|proto]'', tcpdump 可以 指出 那些 捕捉量过小的 数据报, 这里的 proto 是 截断发生处 的 协议层 名称. 注意, 采用 更大的 捕捉范围 既增加了 处理 报文 的 时间, 又 相应的 减少了报文的 缓冲 数量, 可能 导致 报文的丢失. 你 应该 把 snaplen 设的尽量小, 只要 能够 容纳 你 需要 的 协议信息 就可以了.

-T

把 通过 "expression" 挑选出来的 报文 解释成 指定的 type. 目前 已知 的 类型 有: rpc (远程过程调用 Remote Procedure Call), rtp (实时应用协议 Real-Time Applications protocol), rtcp (实时应用控制协议 Real-Time Applications control protocol), vat (可视音频工具 Visual Audio Tool), 和 wb (分布式白板 distributed White Board).

-S

显示 绝对的, 而不是 相对的 TCP 序列号.

-t

禁止 显示 时戳标志.

-tt

显示 未格式化的 时戳标志.

-v

(稍微多一点) 繁琐的输出. 例如, 显示 IP 数据报 中的 生存周期 和 服务类型.

-vv

更繁琐的输出. 例如, 显示 NFS 应答报文 的 附加域.

-w

把 原始报文 存进 file, 而不是 分析 和 显示. 它们 可以 以后 用 -r 选项 显示. 如果 file 是 ``-'', 就 写往 标准输出.

-x

以 16 进制数 形式 显示 每一个 报文 (去掉链路层报头后) . 可以 显示 较小的 完整 报文, 否则 只 显示 snaplen 个 字节 .

expression

用来 选择 要 转储 的 数据报. 如果 没有 指定 expression , 就 转储 网络的 全部 报文. 否则, 只转储 相对 expression 为 `true' 的 数据报.

expression 一个或多个 原语 (primitive) 组成. 原语 通常 由 一个 标识 (id, 名称或数字), 和 标识 前面的 一个或多个 修饰子(qualifier) 组成. 修饰子 有 三种 不同的类型:

type

类型修饰子 指出 标识名称 或 标识数字 代表 什么 类型的东西. 可以使用的 类型 有 host, net 和 port. 例如, `host foo', `net 128.3', `port 20'. 如果 不指定 类型修饰子, 就使用 缺省的 host .

dir

方向修饰子 指出 相对于 标识 的 传输方向 (数据是 传入还是传出 标识). 可以使用的 方向 有 src, dst, src or dst 和 src and dst. 例如, `src foo', `dst net 128.3', `src or dst port ftp-data'. 如果 不指定 方向修饰子, 就使用 缺省的 src or dst . 对于 `null' 链路层 (就是说 象 slip 之类的 点到点 协议), 用 inbound 和 outbound 修饰子 指定 所需的 传输方向.

proto

协议修饰子 要求 匹配 指定的协议. 可以使用的 协议 有: ether, fddi, ip, arp, rarp, decnet, lat, sca, moprc, mopdl, tcp 和 udp. 例如, `ether src foo', `arp net 128.3', `tcp port 21'. 如果 不指定协议修饰子, 就使用 所有 符合 类型 的 协议. 例如, `src foo' 指 `(ip 或 arp 或 rarp) src foo' (注意后者不符合语法), `net bar' 指 `(ip 或 arp 或 rarp) net bar', `port 53' 指 `(tcp 或 udp) port 53'.

[`fddi' 实际上 是 `ether' 的 别名; 分析器 把 它们 视为 ``用在 指定 网络接口 上的 数据链路层.'' FDDI 报头 包含 类似于 以太协议的 源目地址, 而且 通常 包含 类似于 以太协议 的 报文类型, 因此 你 可以过滤 FDDI 域, 就象 分析 以太协议 一样. FDDI 报头 也 包含 其他 域, 但是你 不能 在 过滤器 表达式 里 显式描述.]

作为 上述 的 补充, 有一些 特殊的 `原语' 关键字, 它们 不同于 上面的模式: gateway, broadcast, less, greater 和 数学表达式. 这些 在 后面 有 叙述.

更复杂的 过滤器表达式 可以 通过 and, or 和 not 连接 原语 来 组建. 例如, `host foo and not port ftp and not port ftp-data'. 为了少敲点键, 可以忽略 相同的 修饰子. 例如, `tcp dst port ftp or ftp-data or domain' 实际上 就是 `tcp dst port ftp or tcp dst port ftp-data or tcp dst port domain'.

允许的 原语 有:

dst host host

如果 报文中 IP 的 目的地址域 是 host, 则 逻辑 为 真. host 既可以 是 地址, 也可以 是 主机名.

src host host

如果 报文中 IP 的 源地址域 是 host, 则 逻辑 为 真.

host host

如果 报文中 IP 的 源地址域 或者 目的地址域 是 host, 则 逻辑 为 真. 上面 所有的 host 表达式 都可以 加上 ip, arp, 或 rarp 关键字 做 前缀, 就象:

ip host host

它等价于:

ether proto \ip and host host

如果 host 是 拥有 多个 IP 地址 的 主机名, 它的 每个地址 都会 被查验.

ether dst ehost

如果 报文的 以太目的地址 是 ehost, 则 逻辑 为 真. Ehost 既可以是 名字 (/etc/ethers 里有), 也可以是 数字 (有关 数字格式 另见 ethers(3N) ).

ether src ehost

如果 报文的 以太源地址 是 ehost, 则 逻辑 为 真.

ether host ehost

如果 报文的 以太源地址 或 以太目的地址 是 ehost, 则 逻辑 为 真.

gateway host

如果 报文 把 host 当做 网关, 则 逻辑 为 真. 也就是说, 报文的以太源或目的地址 是 host, 但是 IP 的 源目地址 都不是 host. host 必须 是个 主机名, 而且 必须 存在 /etc/hosts 和 /etc/ethers 中. (一个等价的表达式是

ether host ehost and not host host

对于 host / ehost, 它既可以是 名字, 也可以是 数字.)

dst net net

如果 报文的 IP 目的地址 属于 网络号 net, 则 逻辑 为 真. net 既可以 是 名字 (存在 /etc/networks 中), 也可以是 网络号. (详见 networks(4)).

src net net

如果 报文的 IP 源地址 属于 网络号 net, 则 逻辑 为 真.

net net

如果 报文的 IP 源地址 或 目的地址 属于 网络号 net, 则 逻辑 为 真.

net net mask mask

如果 IP 地址 匹配 指定 网络掩码(netmask) 的 net, 则 逻辑 为 真. 本原语 可以用 src 或 dst 修饰.

net net/len

如果 IP 地址 匹配 指定 网络掩码 的 net, 则 逻辑 为 真, 掩码 的 有效位宽 为 len. 本原语 可以用 src 或 dst 修饰.

dst port port

如果 报文 是 ip/tcp 或 ip/udp, 并且 目的端口 是 port, 则 逻辑 为 真. port 是一个 数字, 也可以是 /etc/services 中 说明过的 名字 (参看 tcp(4P) 和 udp(4P)). 如果 使用 名字, 则 检查 端口号 和 协议. 如果 使用 数字, 或者 有二义的名字, 则 只检查 端口号 (例如, dst port 513 将显示 tcp/login 的数据 和 udp/who 的数据, 而 port domain 将显示 tcp/domain 和 udp/domain 的数据).

src port port

如果 报文 的 源端口号 是 port, 则 逻辑 为 真.

port port

如果 报文 的 源端口 或 目的端口 是 port, 则 逻辑 为 真. 上述的 任意一个 端口表达式 都可以 用 关键字 tcp 或 udp 做 前缀, 就象:

tcp src port port

它 只匹配 源端口 是 port 的 TCP 报文.

less length

如果 报文 的 长度 小于等于 length, 则 逻辑 为 真. 它等同于:

len <= length.

greater length

如果 报文 的 长度 大于等于 length, 则 逻辑 为 真. 它等同于:

len >= length.

ip proto protocol

如果 报文 是 IP 数据报(参见 ip(4P)), 其 内容 的 协议类型 是 protocol, 则 逻辑 为 真. Protocol 可以是 数字, 也可以是 下列 名称 中的 一个: icmp, igrp, udp, nd, 或 tcp. 注意 这些 标识符 tcp, udp, 和 icmp 也同样是 关键字, 所以 必须 用 反斜杠(\) 转义, 在 C-shell 中 应该是 \\ .

ether broadcast

如果 报文 是 以太广播报文, 则 逻辑 为 真. 关键字 ether 是 可选的.

ip broadcast

如果 报文 是 IP广播报文, 则 逻辑 为 真. Tcpdump 检查 全0 和 全1 广播约定, 并且 检查 本地 的 子网掩码.

ether multicast

如果 报文 是 以太多目传送报文(multicast), 则 逻辑 为 真. 关键字 ether 是 可选的. 这实际上 是 `ether[0] & 1 != 0' 的简写.

ip multicast

如果 报文 是 IP多目传送报文, 则 逻辑 为 真.

ether proto protocol

如果 报文协议 属于 以太类型 的 protocol, 则 逻辑 为 真. Protocol 可以是 数字, 也可以是 名字, 如 ip, arp, 或 rarp. 注意 这些 标识符 也是 关键字, 所以 必须 用 反斜杠(\) 转义. [如果是 FDDI (例如, `fddi protocol arp'), 协议 标识 来自 802.2 逻辑链路控制(LLC)报头, 它 通常 位于 FDDI 报头 的 顶层. 当 根据 协议标识过滤 报文 时, Tcpdump 假设 所有的 FDDI 报文 含有 LLC 报头, 而且 LLC 报头 用的是 SNAP 格式.]

decnet src host

如果 DECNET 的 源地址 是 host, 则 逻辑 为 真, 该 主机地址 的 形式 可能 是 ``10.123'', 或者是 DECNET 主机名. [只有 配置成 运行 DECNET 的 Ultrix 系统 支持 DECNET 主机名.]

decnet dst host

如果 DECNET 的 目的地址 是 host, 则 逻辑 为 真.

decnet host host

如果 DECNET 的 源地址 或 目的地址 是 host, 则 逻辑 为 真.

ip, arp, rarp, decnet

是:

ether proto p

的 简写 形式, 其中 p 为 上述 协议 的 一种.

lat, moprc, mopdl

是:

ether proto p

的 简写 形式, 其中 p 为 上述 协议 的 一种. 注意 tcpdump 目前 不知道 如何 分析 这些 协议.

tcp, udp, icmp

是:

ip proto p

的 简写 形式, 其中 p 为 上述 协议 的 一种.

expr relop expr

如果 这个 关系 成立, 则 逻辑 为 真, 其中 relop 是 >, <, >=, <=, =, != 之一, expr 是 数学表达式, 由 常整数(标准C语法形式), 普通的 二进制运算符 [+, -, *, /, &, |], 一个 长度运算符, 和 指定的 报文数据访问算符 组成. 要 访问 报文内 的 数据, 使用 下面的 语法:

proto [ expr : size ]

Proto 是 ether, fddi, ip, arp, rarp, tcp, udp, or icmp 之一, 同时 也指出了 下标 操作 的协议层. expr 给出 字节单位 的 偏移量, 该 偏移量 相对于 指定的 协议层. Size 是 可选项, 指出 感兴趣的 字节数; 它可以 是 1, 2, 4, 缺省为 1 字节. 由 关键字 len 给出的 长度运算符 指明 报文 的 长度.

例如, `ether[0] & 1 != 0' 捕捉 所有的 多目传送 报文. 表达式 `ip[0] & 0xf != 5' 捕捉 所有 带 可选域 的 IP 报文. 表达式 `ip[6:2] & 0x1fff = 0' 只捕捉 未分片 和 片偏移为0 的 数据报. 这种 检查 隐含在 tcp 和 udp 下标操作 中. 例如, tcp[0] 一定是 TCP 报头 的 第一个 字节, 而不是 其中 某个 IP片 的 第一个 字节.

原语 可以 用 下述 方法 结合使用:

园括弧 括起来的 原语 和 操作符 (园括弧 在 Shell 中 有专用, 所以必须转义).

取反操作 (`!' or `not').

连结操作 (`&&' or `and').

或操作 (`||' or `or').

取反操作 有 最高优先级. 或操作 和 连结操作 有 相同的 优先级, 运算时 从左到右 结合. 注意 连结操作 需要 显式的 and 算符, 而不是 并列放置.

如果 给出 标识符, 但没给 关键字, 那么 暗指 最近使用 的 关键字. 例如,

not host vs and ace

作为

not host vs and host ace

的 简写形式, 不应该 和

not ( host vs or ace )

混淆.

表达式参数 可以 作为 单个 参数 传给 tcpdump, 也可以 作为 复合参数, 后者 更方便 一些. 一般说来, 如果 表达式 包含 Shell 元字符(metacharacter), 传递 单个 括起来的 参数 要 容易 一些. 复合参数 在 被解析前 用 空格 联接 一起.

示例 (EXAMPLES)

显示 所有 进出 sundown 的 报文:

tcpdump host sundown

显示 helios 和 主机 hot, ace 之间 的 报文 传送:

tcpdump host helios and \( hot or ace \)

显示 ace 和 除了 helios 以外的 所有 主机 的 IP报文:

tcpdump ip host ace and not helios

显示 本地的主机 和 Berkeley的主机 之间 的 网络数据:

tcpdump net ucb-ether

显示 所有 通过 网关 snup 的 ftp 报文 (注意 这个 表达式 被 单引号 括起, 防止 shell 解释 园括弧):

tcpdump 'gateway snup and (port ftp or ftp-data)'

显示 既不是 来自 本地主机, 也不是 传往 本地主机 的 网络数据 (如果 你 把 网关 通往 某个 其他网络, 这个 做法 将不会 把 数据 发往 你的本地网络).

tcpdump ip and not net localnet

显示 每个 TCP会话 的 起始 和 结束 报文 (SYN 和 FIN 报文), 而且 会话方 中有一个 远程主机.

tcpdump 'tcp[13] & 3 != 0 and not src and dst net localnet'

显示 经过 网关 snup 中 大于 576 字节的 IP 数据报:

tcpdump 'gateway snup and ip[2:2] > 576'

显示 IP 广播 或 多目传送 的 数据报, 这些 报文 不是 通过 以太网 的 广播 或 多目传送 形式 传送的:

tcpdump 'ether[0] & 1 = 0 and ip[16] >= 224'

显示 所有 不是 回响请求/应答 的 ICMP 报文 (也就是说, 不是 ping 报文):

tcpdump 'icmp[0] != 8 and icmp[0] != 0"

输出格式 (OUTPUT FORMAT)

tcpdump 的 输出格式 取决于 协议. 下面的 描述 给出 大多数 格式 的简要说明 和 范例.

链路层报头 (Link Level Headers)

如果 给出 '-e' 选项 就 显示 链路层报头. 在 以太网上, 显示 报文的 源目地址, 协议 和 报文长度.

在 FDDI 网络上, '-e' 选项 导致 tcpdump 显示出 `帧控制(frame control)' 域, 源目地址 和 报文长度. (`帧控制' 域 负责 解释 其余的 报文. 普通报文 (比如说 载有 IP数据报) 是 `异步' 报文, 优先级 介于 0 到 7; 例如, `async4'. 这些 被认为 载有 802.2 逻辑链路控制(LLC) 报文; 如果 它们 不是 ISO 数据报 或者 所谓的 SNAP 报文, 就显示出 LLC 报头.

(注意: 以下 描述中 假设 你 熟悉 RFC-1144 中说明的 SLIP 压缩算法.)

在 SLIP 链路上, tcpdump 显示出 方向指示 (``I'' 指 inbound, ``O'' 指 outbound), 报文类型 和 压缩信息. 首先显示的 是 报文类型. 有三种 类型 ip, utcp 和 ctcp. 对于 ip 报文 不再 显示 更多的 链路信息. 对于 TCP 报文, 在 类型 后面 显示 连接标识. 如果 报文 是 压缩过的, 就显示出 编码的报头. 特殊 情形 以 *S+n 和 *SA+n 的 形式 显示, 这里的 n 是 顺序号 (或顺序号 及其 确认) 发生 的 改变 总和. 如果 不是 特殊 情形, 就显示 0 或 多少个 改变. 改变 由 U (urgent pointer), W (window), A (ack), S (sequence number) 和 I (packet ID) 指明, 后跟 一个 变化量(+n or -n), 或 另一个 值(=n). 最后显示 报文中 的 数据总和, 以及 压缩报头 的 长度.

例如, 下面一行 显示了 一个 传出的 压缩的 TCP 报文, 有一个 隐含的 连接标识; 确认(ack)的 变化量是 6, 顺序号 是 49, 报文ID 是 6; 有三个字节的数据 和六个字节 的 压缩报头:

O ctcp * A+6 S+49 I+6 3 (6)

ARP/RARP 报文

Arp/rarp 报文 的 输出 显示 请求类型 及其 参数. 输出格式 倾向于 能够 自我解释. 这里 是一个 简单的例子, 来自 主机 rtsg 到 主机 csam 的 'rlogin' 开始 部分:

arp who-has csam tell rtsg
arp reply csam is-at CSAM

第一行 说明 rtsg 发出 一个 arp 报文 询问 internet 主机 csam 的 以太网地址. Csam 用 它的 以太地址 作应答 (这个例子中, 以太地址 是 大写的, internet 地址为 小写).

如果 用 tcpdump -n 看上去 要 清楚一些:

arp who-has 128.3.254.6 tell 128.3.254.68
arp reply 128.3.254.6 is-at 02:07:01:00:01:c4

如果 用 tcpdump -e, 可以 看到 实际上 第一个 报文 是 广播, 第二个报文 是 点到点 的:

RTSG Broadcast 0806  64: arp who-has csam tell rtsg
CSAM RTSG 0806  64: arp reply csam is-at CSAM

这里 第一个 报文 指出 以太网源地址是 RTSG, 目的地址 是 以太网广播地址, 类型域 为 16进制数 0806 (类型 ETHER_ARP), 报文全长 64 字节.

TCP 报文

(注意: 以下的描述中 假设 你 熟悉 RFC-793 中 说明的 TCP 协议, 如果 你不了解 这个 协议, 无论是 本文 还是 tcpdump 都对你 用处 不大)

一般说来 tcp 协议的 输出格式是:

src > dst: flags data-seqno ack window urgent options

Src 和 dst 是 源目IP地址和端口. Flags 是 S (SYN), F (FIN), P (PUSH) 或 R (RST) 或 单独的 `.'(无标志), 或者是 它们的 组合. Data-seqno 说明了 本报文中的数据 在 流序号 中的 位置 (见下例). Ack 是 在这条连接上 信源机 希望 下一个 接收的 字节的 流序号 (sequence number). Window 是 在这条连接上 信源机 接收缓冲区 的 字节大小. Urg 表明 报文内 是 `紧急(urgent)' 数据. Options 是 tcp 可选报头, 用 尖括号 括起 (例如, <mss 1024>).

Src, dst 和 flags 肯定 存在. 其他域 依据 报文的 tcp 报头 内容, 只输出 有必要 的 部分.

下面 是 从 主机 rtsg rlogin 到 主机 csam 的 开始部分.

rtsg.1023 > csam.login: S 768512:768512(0) win 4096 <mss 1024>
csam.login > rtsg.1023: S 947648:947648(0) ack 768513 win 4096 <mss 1024>
rtsg.1023 > csam.login: . ack 1 win 4096
rtsg.1023 > csam.login: P 1:2(1) ack 1 win 4096
csam.login > rtsg.1023: . ack 2 win 4096
rtsg.1023 > csam.login: P 2:21(19) ack 1 win 4096
csam.login > rtsg.1023: P 1:2(1) ack 21 win 4077
csam.login > rtsg.1023: P 2:3(1) ack 21 win 4077 urg 1
csam.login > rtsg.1023: P 3:4(1) ack 21 win 4077 urg 1

第一行 是说 从 rtsg 的 tcp 端口 1023 向 csam 的 login 端口 发送 报文. S 标志 表明 设置了 SYN 标志. 报文 的 流序号 是 768512, 没有 数据. (这个写成 `first:last(nbytes)', 意思是 `从 流序号 first 到 last, 不包括 last, 有 nbytes 字节的 用户数据'.) 此时 没有 捎带确认(piggy-backed ack), 有效的 接收窗口 是 4096 字节, 有一个 最大段大小(max-segment-size) 的 选项, 请求 设置 mss 为 1024 字节.

Csam 用类似的 形式 应答, 只是 增加了 一个 对 rtsg SYN 的 捎带确认. 然后 Rtsg 确认 csam 的 SYN. `.' 意味着 没有 设置 标志. 这个 报文 不包含 数据, 因此 也就 没有 数据的流序号. 注意这个 确认流序号 是一个 小整数(1). 当 tcpdump 第一次 发现 一个 tcp 会话时, 它 显示 报文 携带的 流序号. 在 随后收到的 报文里, 它 显示 当前报文 和 最初那个 报文 的 流序号 之 差. 这 意味着 从第一个报文 开始, 以后的 流序号 可以 理解成 数据流 中的 相对位移 as relative byte positions in the conversation's data stream (with the first data byte each direction being `1'). `-S' 选项 能够 改变 这个 特性, 直接 显示 原始的 流序号.

在 第六行, rtsg 传给 csam 19 个字节 的 数据 (字节 2 到 20). 报文中 设置了 PUSH 标志. 第七行 csam 表明 它 收到了 rtsg 的 数据, 字节序号是 21, 但不包括 第21个 字节. 显然 大多数 数据 在 socket 的 缓冲区内, 因为 csam 的 接收窗口 收到的 数据小于 19 个 字节. 同时 csam 向 rtsg 发送了 一个字节 的 数据. 第八和第九行 显示 csam 发送了 两个字节 的 紧急数据 到 rtsg.

如果 捕捉区 设置的 过小, 以至于 tcpdump 不能 捕捉到 完整的 TCP 报头, tcpdump 会 尽可能的 翻译 已捕获的 部分, 然后 显示 ``[|tcp]'', 表明 无法 翻译 其余 部分. 如果 报头 包含 一个 伪造的 选项 (one with a length that's either too small or beyond the end of the header), tcpdump 显示 ``[bad opt]'' 并且 不再 翻译 其他 选项部分 (因为 它 不可能 判断出从哪儿 开始). 如果 报头长度 表明 存在 选项, 但是 IP 数据报 长度 不够, 不可能 真的 保存 选项, tcpdump 就显示 ``[bad hdr length]''.

UDP 报文

UDP 格式 就象 这个 rwho 报文 显示的:

actinide.who > broadcast.who: udp 84

就是说 把一个 udp 数据报 从 主机 actinide 的 who 端口 发送到 broadcast, Internet 广播地址 的 who 端口. 报文 包含 84字节 的 用户数据.

某些 UDP 服务 能够 识别出来(从 源目端口号 上), 因而 显示出 更高层的 协议信息. 特别是 域名服务请求(RFC-1034/1035) 和 NFS 的 RPC 调用(RFC-1050).

UDP 域名服务请求 (Name Server Requests)

(注意: 以下的描述中 假设 你 熟悉 RFC-1035 说明的 域名服务协议. 如果你 不熟悉 这个协议, 下面的内容 就象是 天书.)

域名服务请求 的 格式 是

src > dst: id op? flags qtype qclass name (len)

h2opolo.1538 > helios.domain: 3+ A? ucbvax.berkeley.edu. (37)

主机 h2opolo 访问 helios 上的 域名服务, 询问和 ucbvax.berkeley.edu. 关联的 地址记录(qtype=A). 查询号是 `3'. `+' 表明 设置了 递归请求 标志. 查询长度是 37 字节, 不包括 UDP 和 IP 头. 查询操作 是 普通的 Query 操作, 因此 op 域 可以 忽略. 如果 op 设置成 其他什么东西, 它应该 显示在 `3' 和 `+' 之间. 类似的, qclass 是 普通的 C_IN 类型, 也被 忽略了. 其他类型的 qclass 应该 在 `A' 后面 显示.

Tcpdump 会检查 一些 不规则 情况, 相应的 结果 作为 补充域 放在 方括号内: 如果 某个 查询 包含 回答, 名字服务 或 管理机构部分, 就把 ancount, nscount, 或 arcount 显示成 `[na]', `[nn]' 或 `[nau]', 这里的 n 代表 相应的 数量. 如果 在 第二和第三字节 中, 任何一个 回答位(AA, RA 或 rcode) 或 任何一个 `必须为零' 的位 被 置位, 就显示 `[b2&3=x]', 这里的 x 是 报头 第二和第三字节 的 16进制数.

UDP 名字服务回答

名字服务回答的 格式 是

src > dst:  id op rcode flags a/n/au type class data (len)

helios.domain > h2opolo.1538: 3 3/3/7 A 128.32.137.3 (273)
helios.domain > h2opolo.1537: 2 NXDomain* 0/1/0 (97)

第一个例子里, helios 回答了 h2opolo 发出的 标识为3 的 询问, 一共是 3 个 回答记录, 3 个 名字服务记录 和 7 个管理结构记录. 第一个 回答纪录 的 类型是 A (地址), 数据是 internet 地址 128.32.137.3. 回答的 全长 为 273 字节, 不包括 UDP 和 IP 报头. 作为 A 记录的 class(C_IN) 可以 忽略 op (询问) 和 rcode (NoError).

在第二个例子里, helios 对 标识为2 的 询问 作出 域名不存在 (NXDomain) 的 回答, 没有 回答记录, 一个 名字服务记录, 而且 没有 管理结构.
 `*' 表明 设置了 权威回答(authoritative answer).  由于 没有 回答记录, 这里就 不显示 type, class 和 data.

其他 标志 字符 可以 显示为 `-' (没有设置递归有效(RA)) 和 `|' (设置 消息截短(TC)). 如果 `问题' 部分 没有 有效的 内容, 就 显示 `[nq]'.

注意 名字服务的 询问和回答 一般说来 比较大, 68 字节的 snaplen 可能无法 捕捉到 足够的 报文内容. 如果 你 的确 在 研究 名字服务 的 情况, 可以使用 -s 选项 增大 捕捉缓冲区. `-s 128' 应该 效果 不错了.

NFS 请求和响应

Sun NFS (网络文件系统) 的 请求和响应 显示格式 是:

src.xid > dst.nfs: len op args
src.nfs > dst.xid: reply stat len op results


sushi.6709 > wrl.nfs: 112 readlink fh 21,24/10.73165
wrl.nfs > sushi.6709: reply ok 40 readlink "../var"
sushi.201b > wrl.nfs:
        144 lookup fh 9,74/4096.6878 "xcolors"
wrl.nfs > sushi.201b:
        reply ok 128 lookup fh 9,74/4134.3150

在第一行, 主机 sushi 向 wrl 发送 号码为 6709 的 交易会话 (注意 源主机 后面的 数字 是 交易号, 不是 端口). 这项请求 长 112 字节, 不包括 UDP 和 IP 报头. 在 文件句柄 (fh) 21,24/10.731657119 上执行 readlink (读取 符号连接) 操作. (如果 运气 不错, 就象 这种情况, 文件句柄 可以 依次翻译成 主次设备号, i 节点号, 和 事件号(generation number). ) Wrl 回答 `ok' 和 连接的 内容.

在第三行, sushi 请求 wrl 在 目录文件 9,74/4096.6878 中 查找 `xcolors'. 注意 数据的 打印格式 取决于 操作类型. 格式 应该是 可以自我说明的.

给出 -v (verbose) 选项 可以 显示 附加信息. 例如:

sushi.1372a > wrl.nfs:
        148 read fh 21,11/12.195 8192 bytes @ 24576
wrl.nfs > sushi.1372a:
        reply ok 1472 read REG 100664 ids 417/0 sz 29388

(-v 同时 使它 显示 IP 报头的 TTL, ID, 和 分片域, 在 这个例子里 把它们省略了.) 在第一行, sushi 请求 wrl 从 文件 21,11/12.195 的 偏移位置 24576 开始, 读取 8192 字节. Wrl 回答 `ok'; 第二行 显示的 报文 是 应答的 第一个 分片, 因此 只有 1472 字节 (其余数据 在 后续的 分片中传过来, 但由于 这些分片里 没有 NFS 甚至 UDP 报头, 因此 根据 所使用的 过滤器表达式, 有可能 不显示). -v 选项 还会 显示 一些 文件属性 (它们 作为 文件数据 的 附带部分 传回来): 文件类型 (普通文件 ``REG''), 存取模式 (八进制数), uid 和 gid, 以及 文件大小.

如果再给一个 -v 选项 (-vv), 还能 显示 更多的细节.

注意 NFS 请求 的 数据量 非常大, 除非 增加 snaplen, 否则 很多细节 无法显示. 试一试 `-s 192' 选项.

NFS 应答报文 没有明确 标明 RPC 操作. 因此 tcpdump 保留有 ``近来的'' 请求 记录, 根据 交易号 匹配 应答报文. 如果 应答报文 没有 相应的 请求报文, 它 就 无法分析.

KIP Appletalk (UDP 上的 DDP)

Appletalk DDP 报文 封装在 UDP 数据报 中, 解包后 按 DDP 报文 转储 (也就是说, 忽略 所有的 UDP 报头 信息). 文件 /etc/atalk.names 用来 把 appletalk 网络和节点号 翻译成 名字. 这个文件 的 行格式 是

number  name

1.254           ether
16.1            icsd-net
1.254.110       ace

前两行 给出了 appletalk 的 网络名称. 第三行 给出 某个主机 的 名字 (主机和网络 依据 第三组 数字 区分 - 网络号 一定 是 两组数字, 主机号 一定 是 三组 数字.) 号码 和 名字 用 空白符(空格或tab) 隔开. /etc/atalk.names 文件 可以 包含 空行 或 注释行(以`#'开始的行).

Appletalk 地址 按 这个格式 显示

net.host.port

144.1.209.2 > icsd-net.112.220
office.2 > icsd-net.112.220
jssmag.149.235 > icsd-net.2

(如果 不存在 /etc/atalk.names , 或者 里面 缺少 有效项目, 就以 数字形式 显示 地址.) 第一个例子里, 网络 144.1 的 209 节点的 NBP (DDP 端口 2) 向 网络 icsd 的 112 节点 的 220 端口 发送数据. 第二行 和 上面 一样, 只是 知道了 源节点 的 全称 (`office'). 第三行 是从 网络 jssmag 的 149 节点 的 235 端口 向 icsd-net 的 NBP 端口广播 (注意 广播地址 (255) 隐含在 无主机号的 网络名字 中 - 所以 在 /etc/atalk.names 中 区分 节点名 和 网络名 是个 好主意).

Tcpdump 可以 翻译 NBP (名字联结协议) 和 ATP (Appletalk 交互协议) 的 报文内容. 其他协议 只转储 协议名称 (或号码, 如果 还 没给 这个协议 注册 名称) 和 报文大小.

NBP 报文 的 输出格式 就象 下面的 例子:

icsd-net.112.220 > jssmag.2: nbp-lkup 190: "=:LaserWriter@*"
jssmag.209.2 > icsd-net.112.220: nbp-reply 190: "RM1140:LaserWriter@*" 250
techpit.2 > icsd-net.112.220: nbp-reply 190: "techpit:LaserWriter@*" 186

第一行 是 网络 icsd 的 112 主机 在 网络 jssmag 上的 广播, 对 名字 laserwriter 做 名字查询请求. 名字查询请求 的 nbp 标识号 是 190. 第二行 显示的是 对 这个请求 的 回答 (注意 它们 有 同样的 标识号), 主机 jssmag.209 表示 在它的 250 端口 注册了 一个 laserwriter 的 资源, 名字是 "RM1140". 第三行 是 这个请求 的 其他回答, 主机 techpit 的 186 端口 有 laserwriter 注册的 "techpit".

ATP 报文 格式 如 下例 所示:

jssmag.209.165 > helios.132: atp-req  12266<0-7> 0xae030001
helios.132 > jssmag.209.165: atp-resp 12266:0 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:1 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:2 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:4 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:6 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp*12266:7 (512) 0xae040000
jssmag.209.165 > helios.132: atp-req  12266<3,5> 0xae030001
helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
jssmag.209.165 > helios.132: atp-rel  12266<0-7> 0xae030001
jssmag.209.133 > helios.132: atp-req* 12267<0-7> 0xae030002

Jssmag.209 向 主机 helios 发起 12266 号 交易, 请求 8 个 报文(`<0-7>'). 行尾的 十六进制数 是 请求中 `userdata' 域 的 值.

Helios 用 8 个 512字节 的 报文 应答. 跟在 交易号 后面的 `:digit' 给出了 交易过程中 报文的 序列号, 括弧内的 数字 是 报文的 数据量, 不包括 atp 报头. 报文 7 的 `*' 表明 设置了 EOM 位.

然后 Jssmag.209 请求 重传 第 3 & 5 报文. Helios 做了 重传后 jssmag.209 结束 这次 交易. 最后, jssmag.209 发起 下一次 交易请求. 请求中的 `*' 表明 没有 设置 XO (只有一次) 位.

IP 分片

分片的 Internet 数据报 显示为

(frag id:size@offset+)
(frag id:size@offset)

(第一种 形式 表明 还有 更多的 分片. 第二种 形式 表明 这是 最后 一片.)

Id 是 分片 标识号. Size 是 分片 大小 (字节), 不包括 IP 报头. Offset 是 该分片 在 原数据报 中 的 偏移 (单位是字节).

每一个 分片 的 信息 都可以 打印出来. 第一个 分片 包含了 高层 协议 报头, 显示 协议信息 后 显示 分片 的 信息. 第一个 分片 以后的 分片 不再 含有高层协议 报头, 所以 在 源目地址 后面 只显示 分片 信息. 例如, 下面是 从 arizona.edu 到 lbl-rtsg.arpa 的 一部分 ftp 传输, 途经的 CSNET 看上去 处理不了 576 字节的 数据报:

arizona.ftp-data > rtsg.1170: . 1024:1332(308) ack 1 win 4096 (frag 595a:328@0+)
arizona > rtsg: (frag 595a:204@328)
rtsg.1170 > arizona.ftp-data: . ack 1536 win 2560

这里 有几点 需要注意: 首先, 第二行的 地址 不包括 端口号. 这是因为 TCP 协议 信息 全部 装到了 第一个 分片内, 所以 显示 后续分片的 时候 不可能 知道端口 或 流序号. 其次, 第一行的 tcp 流序号部分 看上去有 308 字节的 用户数据, 实际上 是 512 字节 (第一个 分片的 308 和 第二个 分片的 204 字节). 如果你 正在 寻找 流序号中 的 空洞, 或者 试图 匹配 报文 的 确认(ack), 那你上当了.

如果 报文的 IP 标有 不要分片 标志, 显示时 在尾部 加上 (DF).

时戳

缺省情况下, 所有 输出行 的 前面 都有 时戳. 时戳 就是 当前时间, 显示格式为

hh:mm:ss.frac

精度 和 内核时钟 一样. 时戳 反映了 内核 收到 报文 的 时间. 从 以太接口 收到 报文 到 内核 响应 '报文就绪' 中断 有一个 滞后, 该 滞后 不被考虑.

另见 (SEE ALSO)

traffic(1C), nit(4P), bpf(4), pcap(3)

#p#

NAME

tcpdump - dump traffic on a network  

SYNOPSIS

tcpdump [ -AdDeflLnNOpqRStuUvxX ] [ -c count ]

         [ -C file_size ] [ -F file ]

         [ -i interface ] [ -m module ] [ -M secret ] [ -r file ]

         [ -s snaplen ] [ -T type ] [ -w file ]

         [ -W filecount ] [ -E spi@ipaddr algo:secret,... ]

         [ -y datalinktype ] [ -Z user ]
         [ expression ]
 

DESCRIPTION

Tcpdump prints out the headers of packets on a network interface that match the boolean expression. It can also be run with the -w flag, which causes it to save the packet data to a file for later analysis, and/or with the -r flag, which causes it to read from a saved packet file rather than to read packets from a network interface. In all cases, only packets that match expression will be processed by tcpdump.

Tcpdump will, if not run with the -c flag, continue capturing packets until it is interrupted by a SIGINT signal (generated, for example, by typing your interrupt character, typically control-C) or a SIGTERM signal (typically generated with the kill(1) command); if run with the -c flag, it will capture packets until it is interrupted by a SIGINT or SIGTERM signal or the specified number of packets have been processed.

When tcpdump finishes capturing packets, it will report counts of:

packets ``captured'' (this is the number of packets that tcpdump has received and processed);

packets ``received by filter'' (the meaning of this depends on the OS on which you're running tcpdump, and possibly on the way the OS was configured - if a filter was specified on the command line, on some OSes it counts packets regardless of whether they were matched by the filter expression and, even if they were matched by the filter expression, regardless of whether tcpdump has read and processed them yet, on other OSes it counts only packets that were matched by the filter expression regardless of whether tcpdump has read and processed them yet, and on other OSes it counts only packets that were matched by the filter expression and were processed by tcpdump);

packets ``dropped by kernel'' (this is the number of packets that were dropped, due to a lack of buffer space, by the packet capture mechanism in the OS on which tcpdump is running, if the OS reports that information to applications; if not, it will be reported as 0).

On platforms that support the SIGINFO signal, such as most BSDs (including Mac OS X) and Digital/Tru64 UNIX, it will report those counts when it receives a SIGINFO signal (generated, for example, by typing your ``status'' character, typically control-T, although on some platforms, such as Mac OS X, the ``status'' character is not set by default, so you must set it with stty(1) in order to use it) and will continue capturing packets.

Reading packets from a network interface may require that you have special privileges:

Under SunOS 3.x or 4.x with NIT or BPF:

You must have read access to /dev/nit or /dev/bpf*.

Under Solaris with DLPI:

You must have read/write access to the network pseudo device, e.g. /dev/le. On at least some versions of Solaris, however, this is not sufficient to allow tcpdump to capture in promiscuous mode; on those versions of Solaris, you must be root, or tcpdump must be installed setuid to root, in order to capture in promiscuous mode. Note that, on many (perhaps all) interfaces, if you don't capture in promiscuous mode, you will not see any outgoing packets, so a capture not done in promiscuous mode may not be very useful.

Under HP-UX with DLPI:

You must be root or tcpdump must be installed setuid to root.

Under IRIX with snoop:

You must be root or tcpdump must be installed setuid to root.

Under Linux:

You must be root or tcpdump must be installed setuid to root (unless your distribution has a kernel that supports capability bits such as CAP_NET_RAW and code to allow those capability bits to be given to particular accounts and to cause those bits to be set on a user's initial processes when they log in, in which case you must have CAP_NET_RAW in order to capture and CAP_NET_ADMIN to enumerate network devices with, for example, the -D flag).

Under ULTRIX and Digital UNIX/Tru64 UNIX:

Any user may capture network traffic with tcpdump. However, no user (not even the super-user) can capture in promiscuous mode on an interface unless the super-user has enabled promiscuous-mode operation on that interface using pfconfig(8), and no user (not even the super-user) can capture unicast traffic received by or sent by the machine on an interface unless the super-user has enabled copy-all-mode operation on that interface using pfconfig, so useful packet capture on an interface probably requires that either promiscuous-mode or copy-all-mode operation, or both modes of operation, be enabled on that interface.

Under BSD (this includes Mac OS X):

You must have read access to /dev/bpf*. On BSDs with a devfs (this includes Mac OS X), this might involve more than just having somebody with super-user access setting the ownership or permissions on the BPF devices - it might involve configuring devfs to set the ownership or permissions every time the system is booted, if the system even supports that; if it doesn't support that, you might have to find some other way to make that happen at boot time.

Reading a saved packet file doesn't require special privileges.  

OPTIONS

-A

Print each packet (minus its link level header) in ASCII. Handy for capturing web pages.

-c

Exit after receiving count packets.

-C

Before writing a raw packet to a savefile, check whether the file is currently larger than file_size and, if so, close the current savefile and open a new one. Savefiles after the first savefile will have the name specified with the -w flag, with a number after it, starting at 1 and continuing upward. The units of file_size are millions of bytes (1,000,000 bytes, not 1,048,576 bytes).

-d

Dump the compiled packet-matching code in a human readable form to standard output and stop.

-dd

Dump packet-matching code as a C program fragment.

-ddd

Dump packet-matching code as decimal numbers (preceded with a count).

-D

Print the list of the network interfaces available on the system and on which tcpdump can capture packets. For each network interface, a number and an interface name, possibly followed by a text description of the interface, is printed. The interface name or the number can be supplied to the -i flag to specify an interface on which to capture.

This can be useful on systems that don't have a command to list them (e.g., Windows systems, or UNIX systems lacking ifconfig -a); the number can be useful on Windows 2000 and later systems, where the interface name is a somewhat complex string.

The -D flag will not be supported if tcpdump was built with an older version of libpcap that lacks the pcap_findalldevs() function.

-e

Print the link-level header on each dump line.

-E

Use spi@ipaddr algo:secret for decrypting IPsec ESP packets that are addressed to addr and contain Security Parameter Index value spi. This combination may be repeated with comma or newline seperation.

Note that setting the secret for IPv4 ESP packets is supported at this time.

Algorithms may be des-cbc, 3des-cbc, blowfish-cbc, rc3-cbc, cast128-cbc, or none. The default is des-cbc. The ability to decrypt packets is only present if tcpdump was compiled with cryptography enabled.

secret is the ASCII text for ESP secret key. If preceeded by 0x, then a hex value will be read.

The option assumes RFC2406 ESP, not RFC1827 ESP. The option is only for debugging purposes, and the use of this option with a true `secret' key is discouraged. By presenting IPsec secret key onto command line you make it visible to others, via ps(1) and other occasions.

In addition to the above syntax, the syntax file name may be used to have tcpdump read the provided file in. The file is opened upon receiving the first ESP packet, so any special permissions that tcpdump may have been given should already have been given up.

-f

Print `foreign' IPv4 addresses numerically rather than symbolically (this option is intended to get around serious brain damage in Sun's NIS server --- usually it hangs forever translating non-local internet numbers).

The test for `foreign' IPv4 addresses is done using the IPv4 address and netmask of the interface on which capture is being done. If that address or netmask are not available, available, either because the interface on which capture is being done has no address or netmask or because the capture is being done on the Linux "any" interface, which can capture on more than one interface, this option will not work correctly.

-F

Use file as input for the filter expression. An additional expression given on the command line is ignored.

-i

Listen on interface. If unspecified, tcpdump searches the system interface list for the lowest numbered, configured up interface (excluding loopback). Ties are broken by choosing the earliest match.

On Linux systems with 2.2 or later kernels, an interface argument of ``any'' can be used to capture packets from all interfaces. Note that captures on the ``any'' device will not be done in promiscuous mode.

If the -D flag is supported, an interface number as printed by that flag can be used as the interface argument.

-l

Make stdout line buffered. Useful if you want to see the data while capturing it. E.g.,
``tcpdump  -l  |  tee dat'' or ``tcpdump  -l   > dat  &  tail  -f  dat''.

-L

List the known data link types for the interface and exit.

-m

Load SMI MIB module definitions from file module. This option can be used several times to load several MIB modules into tcpdump.

-M

Use secret as a shared secret for validating the digests found in TCP segments with the TCP-MD5 option (RFC 2385), if present.

-n

Don't convert host addresses to names. This can be used to avoid DNS lookups.

-nn

Don't convert protocol and port numbers etc. to names either.

-N

Don't print domain name qualification of host names. E.g., if you give this flag then tcpdump will print ``nic'' instead of ``nic.ddn.mil''.

-O

Do not run the packet-matching code optimizer. This is useful only if you suspect a bug in the optimizer.

-p

Don't put the interface into promiscuous mode. Note that the interface might be in promiscuous mode for some other reason; hence, `-p' cannot be used as an abbreviation for `ether host {local-hw-addr} or ether broadcast'.

-q

Quick (quiet?) output. Print less protocol information so output lines are shorter.

-R

Assume ESP/AH packets to be based on old specification (RFC1825 to RFC1829). If specified, tcpdump will not print replay prevention field. Since there is no protocol version field in ESP/AH specification, tcpdump cannot deduce the version of ESP/AH protocol.

-r

Read packets from file (which was created with the -w option). Standard input is used if file is ``-''.

-S

Print absolute, rather than relative, TCP sequence numbers.

-s

Snarf snaplen bytes of data from each packet rather than the default of 68 (with SunOS's NIT, the minimum is actually 96). 68 bytes is adequate for IP, ICMP, TCP and UDP but may truncate protocol information from name server and NFS packets (see below). Packets truncated because of a limited snapshot are indicated in the output with ``[|proto]'', where proto is the name of the protocol level at which the truncation has occurred. Note that taking larger snapshots both increases the amount of time it takes to process packets and, effectively, decreases the amount of packet buffering. This may cause packets to be lost. You should limit snaplen to the smallest number that will capture the protocol information you're interested in. Setting snaplen to 0 means use the required length to catch whole packets.

-T

Force packets selected by "expression" to be interpreted the specified type. Currently known types are aodv (Ad-hoc On-demand Distance Vector protocol), cnfp (Cisco NetFlow protocol), rpc (Remote Procedure Call), rtp (Real-Time Applications protocol), rtcp (Real-Time Applications control protocol), snmp (Simple Network Management Protocol), tftp (Trivial File Transfer Protocol), vat (Visual Audio Tool), and wb (distributed White Board).

-t

Don't print a timestamp on each dump line.

-tt

Print an unformatted timestamp on each dump line.

-ttt

Print a delta (in micro-seconds) between current and previous line on each dump line.

-tttt

Print a timestamp in default format proceeded by date on each dump line.

-u

Print undecoded NFS handles.

-U

Make output saved via the -w option ``packet-buffered''; i.e., as each packet is saved, it will be written to the output file, rather than being written only when the output buffer fills.

The -U flag will not be supported if tcpdump was built with an older version of libpcap that lacks the pcap_dump_flush() function.

-v

When parsing and printing, produce (slightly more) verbose output. For example, the time to live, identification, total length and options in an IP packet are printed. Also enables additional packet integrity checks such as verifying the IP and ICMP header checksum.

When writing to a file with the -w option, report, every 10 seconds, the number of packets captured.

-vv

Even more verbose output. For example, additional fields are printed from NFS reply packets, and SMB packets are fully decoded.

-vvv

Even more verbose output. For example, telnet SB ... SE options are printed in full. With -X Telnet options are printed in hex as well.

-w

Write the raw packets to file rather than parsing and printing them out. They can later be printed with the -r option. Standard output is used if file is ``-''.

-W

Used in conjunction with the -C option, this will limit the number of files created to the specified number, and begin overwriting files from the beginning, thus creating a 'rotating' buffer. In addition, it will name the files with enough leading 0s to support the maximum number of files, allowing them to sort correctly.

-x

Print each packet (minus its link level header) in hex. The smaller of the entire packet or snaplen bytes will be printed. Note that this is the entire link-layer packet, so for link layers that pad (e.g. Ethernet), the padding bytes will also be printed when the higher layer packet is shorter than the required padding.

-xx

Print each packet, including its link level header, in hex.

-X

Print each packet (minus its link level header) in hex and ASCII. This is very handy for analysing new protocols.

-XX

Print each packet, including its link level header, in hex and ASCII.

-y

Set the data link type to use while capturing packets to datalinktype.

-Z

Drops privileges (if root) and changes user ID to user and the group ID to the primary group of user.

This behavior can also be enabled by default at compile time.

expression

selects which packets will be dumped. If no expression is given, all packets on the net will be dumped. Otherwise, only packets for which expression is `true' will be dumped.

The expression consists of one or more primitives. Primitives usually consist of an id (name or number) preceded by one or more qualifiers. There are three different kinds of qualifier:

type

qualifiers say what kind of thing the id name or number refers to. Possible types are host, net and port. E.g., `host foo', `net 128.3', `port 20'. If there is no type qualifier, host is assumed.

dir

qualifiers specify a particular transfer direction to and/or from id. Possible directions are src, dst, src or dst and src and dst. E.g., `src foo', `dst net 128.3', `src or dst port ftp-data'. If there is no dir qualifier, src or dst is assumed. For some link layers, such as SLIP and the ``cooked'' Linux capture mode used for the ``any'' device and for some other device types, the inbound and outbound qualifiers can be used to specify a desired direction.

proto

qualifiers restrict the match to a particular protocol. Possible protos are: ether, fddi, tr, wlan, ip, ip6, arp, rarp, decnet, tcp and udp. E.g., `ether src foo', `arp net 128.3', `tcp port 21'. If there is no proto qualifier, all protocols consistent with the type are assumed. E.g., `src foo' means `(ip or arp or rarp) src foo' (except the latter is not legal syntax), `net bar' means `(ip or arp or rarp) net bar' and `port 53' means `(tcp or udp) port 53'.

[`fddi' is actually an alias for `ether'; the parser treats them identically as meaning ``the data link level used on the specified network interface.'' FDDI headers contain Ethernet-like source and destination addresses, and often contain Ethernet-like packet types, so you can filter on these FDDI fields just as with the analogous Ethernet fields. FDDI headers also contain other fields, but you cannot name them explicitly in a filter expression.

Similarly, `tr' and `wlan' are aliases for `ether'; the previous paragraph's statements about FDDI headers also apply to Token Ring and 802.11 wireless LAN headers. For 802.11 headers, the destination address is the DA field and the source address is the SA field; the BSSID, RA, and TA fields aren't tested.]

In addition to the above, there are some special `primitive' keywords that don't follow the pattern: gateway, broadcast, less, greater and arithmetic expressions. All of these are described below.

More complex filter expressions are built up by using the words and, or and not to combine primitives. E.g., `host foo and not port ftp and not port ftp-data'. To save typing, identical qualifier lists can be omitted. E.g., `tcp dst port ftp or ftp-data or domain' is exactly the same as `tcp dst port ftp or tcp dst port ftp-data or tcp dst port domain'.

Allowable primitives are:

dst host host

True if the IPv4/v6 destination field of the packet is host, which may be either an address or a name.

src host host

True if the IPv4/v6 source field of the packet is host.

host host

True if either the IPv4/v6 source or destination of the packet is host. Any of the above host expressions can be prepended with the keywords, ip, arp, rarp, or ip6 as in:

ip host host

which is equivalent to:

ether proto \ip and host host

If host is a name with multiple IP addresses, each address will be checked for a match.

ether dst ehost

True if the ethernet destination address is ehost. Ehost may be either a name from /etc/ethers or a number (see ethers(3N) for numeric format).

ether src ehost

True if the ethernet source address is ehost.

ether host ehost

True if either the ethernet source or destination address is ehost.

gateway host

True if the packet used host as a gateway. I.e., the ethernet source or destination address was host but neither the IP source nor the IP destination was host. Host must be a name and must be found both by the machine's host-name-to-IP-address resolution mechanisms (host name file, DNS, NIS, etc.) and by the machine's host-name-to-Ethernet-address resolution mechanism (/etc/ethers, etc.). (An equivalent expression is

ether host ehost and not host host

which can be used with either names or numbers for host / ehost.) This syntax does not work in IPv6-enabled configuration at this moment.

dst net net

True if the IPv4/v6 destination address of the packet has a network number of net. Net may be either a name from /etc/networks or a network number (see networks(4) for details).

src net net

True if the IPv4/v6 source address of the packet has a network number of net.

net net

True if either the IPv4/v6 source or destination address of the packet has a network number of net.

net net mask netmask

True if the IP address matches net with the specific netmask. May be qualified with src or dst. Note that this syntax is not valid for IPv6 net.

net net/len

True if the IPv4/v6 address matches net with a netmask len bits wide. May be qualified with src or dst.

dst port port

True if the packet is ip/tcp, ip/udp, ip6/tcp or ip6/udp and has a destination port value of port. The port can be a number or a name used in /etc/services (see tcp(4P) and udp(4P)). If a name is used, both the port number and protocol are checked. If a number or ambiguous name is used, only the port number is checked (e.g., dst port 513 will print both tcp/login traffic and udp/who traffic, and port domain will print both tcp/domain and udp/domain traffic).

src port port

True if the packet has a source port value of port.

port port

True if either the source or destination port of the packet is port. Any of the above port expressions can be prepended with the keywords, tcp or udp, as in:

tcp src port port

which matches only tcp packets whose source port is port.

less length

True if the packet has a length less than or equal to length. This is equivalent to:

len <= length.

greater length

True if the packet has a length greater than or equal to length. This is equivalent to:

len >= length.

ip proto protocol

True if the packet is an IP packet (see ip(4P)) of protocol type protocol. Protocol can be a number or one of the names icmp, icmp6, igmp, igrp, pim, ah, esp, vrrp, udp, or tcp. Note that the identifiers tcp, udp, and icmp are also keywords and must be escaped via backslash (\), which is \\ in the C-shell. Note that this primitive does not chase the protocol header chain.

ip6 proto protocol

True if the packet is an IPv6 packet of protocol type protocol. Note that this primitive does not chase the protocol header chain.

ip6 protochain protocol

True if the packet is IPv6 packet, and contains protocol header with type protocol in its protocol header chain. For example,

ip6 protochain 6

matches any IPv6 packet with TCP protocol header in the protocol header chain. The packet may contain, for example, authentication header, routing header, or hop-by-hop option header, between IPv6 header and TCP header. The BPF code emitted by this primitive is complex and cannot be optimized by BPF optimizer code in tcpdump, so this can be somewhat slow.

ip protochain protocol

Equivalent to ip6 protochain protocol, but this is for IPv4.

ether broadcast

True if the packet is an ethernet broadcast packet. The ether keyword is optional.

ip broadcast

True if the packet is an IPv4 broadcast packet. It checks for both the all-zeroes and all-ones broadcast conventions, and looks up the subnet mask on the interface on which the capture is being done.

If the subnet mask of the interface on which the capture is being done is not available, either because the interface on which capture is being done has no netmask or because the capture is being done on the Linux "any" interface, which can capture on more than one interface, this check will not work correctly.

ether multicast

True if the packet is an ethernet multicast packet. The ether keyword is optional. This is shorthand for `ether[0] & 1 != 0'.

ip multicast

True if the packet is an IP multicast packet.

ip6 multicast

True if the packet is an IPv6 multicast packet.

ether proto protocol

True if the packet is of ether type protocol. Protocol can be a number or one of the names ip, ip6, arp, rarp, atalk, aarp, decnet, sca, lat, mopdl, moprc, iso, stp, ipx, or netbeui. Note these identifiers are also keywords and must be escaped via backslash (\).

[In the case of FDDI (e.g., `fddi protocol arp'), Token Ring (e.g., `tr protocol arp'), and IEEE 802.11 wireless LANS (e.g., `wlan protocol arp'), for most of those protocols, the protocol identification comes from the 802.2 Logical Link Control (LLC) header, which is usually layered on top of the FDDI, Token Ring, or 802.11 header.

When filtering for most protocol identifiers on FDDI, Token Ring, or 802.11, tcpdump checks only the protocol ID field of an LLC header in so-called SNAP format with an Organizational Unit Identifier (OUI) of 0x000000, for encapsulated Ethernet; it doesn't check whether the packet is in SNAP format with an OUI of 0x000000. The exceptions are:

iso

tcpdump checks the DSAP (Destination Service Access Point) and SSAP (Source Service Access Point) fields of the LLC header;

stp and netbeui

tcpdump checks the DSAP of the LLC header;

atalk

tcpdump checks for a SNAP-format packet with an OUI of 0x080007 and the AppleTalk etype.

In the case of Ethernet, tcpdump checks the Ethernet type field for most of those protocols. The exceptions are:

iso, sap, and netbeui

tcpdump checks for an 802.3 frame and then checks the LLC header as it does for FDDI, Token Ring, and 802.11;

atalk

tcpdump checks both for the AppleTalk etype in an Ethernet frame and for a SNAP-format packet as it does for FDDI, Token Ring, and 802.11;

aarp

tcpdump checks for the AppleTalk ARP etype in either an Ethernet frame or an 802.2 SNAP frame with an OUI of 0x000000;

ipx

tcpdump checks for the IPX etype in an Ethernet frame, the IPX DSAP in the LLC header, the 802.3-with-no-LLC-header encapsulation of IPX, and the IPX etype in a SNAP frame.

decnet src host

True if the DECNET source address is host, which may be an address of the form ``10.123'', or a DECNET host name. [DECNET host name support is only available on ULTRIX systems that are configured to run DECNET.]

decnet dst host

True if the DECNET destination address is host.

decnet host host

True if either the DECNET source or destination address is host.

ifname interface

True if the packet was logged as coming from the specified interface (applies only to packets logged by OpenBSD's pf(4)).

on interface

Synonymous with the ifname modifier.

rnr num

True if the packet was logged as matching the specified PF rule number (applies only to packets logged by OpenBSD's pf(4)).

rulenum num

Synonomous with the rnr modifier.

reason code

True if the packet was logged with the specified PF reason code. The known codes are: match, bad-offset, fragment, short, normalize, and memory (applies only to packets logged by OpenBSD's pf(4)).

rset name

True if the packet was logged as matching the specified PF ruleset name of an anchored ruleset (applies only to packets logged by pf(4)).

ruleset name

Synonomous with the rset modifier.

srnr num

True if the packet was logged as matching the specified PF rule number of an anchored ruleset (applies only to packets logged by pf(4)).

subrulenum num

Synonomous with the srnr modifier.

action act

True if PF took the specified action when the packet was logged. Known actions are: pass and block (applies only to packets logged by OpenBSD's pf(4)).

ip, ip6, arp, rarp, atalk, aarp, decnet, iso, stp, ipx, netbeui

Abbreviations for:

ether proto p

where p is one of the above protocols.

lat, moprc, mopdl

Abbreviations for:

ether proto p

where p is one of the above protocols. Note that tcpdump does not currently know how to parse these protocols.

vlan [vlan_id]

True if the packet is an IEEE 802.1Q VLAN packet. If [vlan_id] is specified, only true is the packet has the specified vlan_id. Note that the first vlan keyword encountered in expression changes the decoding offsets for the remainder of expression on the assumption that the packet is a VLAN packet.

tcp, udp, icmp

Abbreviations for:

ip proto p or ip6 proto p

where p is one of the above protocols.

iso proto protocol

True if the packet is an OSI packet of protocol type protocol. Protocol can be a number or one of the names clnp, esis, or isis.

clnp, esis, isis

Abbreviations for:

iso proto p

where p is one of the above protocols.

l1, l2, iih, lsp, snp, csnp, psnp

Abbreviations for IS-IS PDU types.

vpi n

True if the packet is an ATM packet, for SunATM on Solaris, with a virtual path identifier of n.

vci n

True if the packet is an ATM packet, for SunATM on Solaris, with a virtual channel identifier of n.

lane

True if the packet is an ATM packet, for SunATM on Solaris, and is an ATM LANE packet. Note that the first lane keyword encountered in expression changes the tests done in the remainder of expression on the assumption that the packet is either a LANE emulated Ethernet packet or a LANE LE Control packet. If lane isn't specified, the tests are done under the assumption that the packet is an LLC-encapsulated packet.

llc

True if the packet is an ATM packet, for SunATM on Solaris, and is an LLC-encapsulated packet.

oamf4s

True if the packet is an ATM packet, for SunATM on Solaris, and is a segment OAM F4 flow cell (VPI=0 & VCI=3).

oamf4e

True if the packet is an ATM packet, for SunATM on Solaris, and is an end-to-end OAM F4 flow cell (VPI=0 & VCI=4).

oamf4

True if the packet is an ATM packet, for SunATM on Solaris, and is a segment or end-to-end OAM F4 flow cell (VPI=0 & (VCI=3 | VCI=4)).

oam

True if the packet is an ATM packet, for SunATM on Solaris, and is a segment or end-to-end OAM F4 flow cell (VPI=0 & (VCI=3 | VCI=4)).

metac

True if the packet is an ATM packet, for SunATM on Solaris, and is on a meta signaling circuit (VPI=0 & VCI=1).

bcc

True if the packet is an ATM packet, for SunATM on Solaris, and is on a broadcast signaling circuit (VPI=0 & VCI=2).

sc

True if the packet is an ATM packet, for SunATM on Solaris, and is on a signaling circuit (VPI=0 & VCI=5).

ilmic

True if the packet is an ATM packet, for SunATM on Solaris, and is on an ILMI circuit (VPI=0 & VCI=16).

connectmsg

True if the packet is an ATM packet, for SunATM on Solaris, and is on a signaling circuit and is a Q.2931 Setup, Call Proceeding, Connect, Connect Ack, Release, or Release Done message.

metaconnect

True if the packet is an ATM packet, for SunATM on Solaris, and is on a meta signaling circuit and is a Q.2931 Setup, Call Proceeding, Connect, Release, or Release Done message.

expr relop expr

True if the relation holds, where relop is one of >, <, >=, <=, =, !=, and expr is an arithmetic expression composed of integer constants (expressed in standard C syntax), the normal binary operators [+, -, *, /, &, |, <<, >>], a length operator, and special packet data accessors. To access data inside the packet, use the following syntax:

proto [ expr : size ]

Proto is one of ether, fddi, tr, wlan, ppp, slip, link, ip, arp, rarp, tcp, udp, icmp or ip6, and indicates the protocol layer for the index operation. (ether, fddi, wlan, tr, ppp, slip and link all refer to the link layer.) Note that tcp, udp and other upper-layer protocol types only apply to IPv4, not IPv6 (this will be fixed in the future). The byte offset, relative to the indicated protocol layer, is given by expr. Size is optional and indicates the number of bytes in the field of interest; it can be either one, two, or four, and defaults to one. The length operator, indicated by the keyword len, gives the length of the packet.

For example, `ether[0] & 1 != 0' catches all multicast traffic. The expression `ip[0] & 0xf != 5' catches all IP packets with options. The expression `ip[6:2] & 0x1fff = 0' catches only unfragmented datagrams and frag zero of fragmented datagrams. This check is implicitly applied to the tcp and udp index operations. For instance, tcp[0] always means the first byte of the TCP header, and never means the first byte of an intervening fragment.

Some offsets and field values may be expressed as names rather than as numeric values. The following protocol header field offsets are available: icmptype (ICMP type field), icmpcode (ICMP code field), and tcpflags (TCP flags field).

The following ICMP type field values are available: icmp-echoreply, icmp-unreach, icmp-sourcequench, icmp-redirect, icmp-echo, icmp-routeradvert, icmp-routersolicit, icmp-timxceed, icmp-paramprob, icmp-tstamp, icmp-tstampreply, icmp-ireq, icmp-ireqreply, icmp-maskreq, icmp-maskreply.

The following TCP flags field values are available: tcp-fin, tcp-syn, tcp-rst, tcp-push, tcp-ack, tcp-urg.

Primitives may be combined using:

A parenthesized group of primitives and operators (parentheses are special to the Shell and must be escaped).

Negation (`!' or `not').

Concatenation (`&&' or `and').

Alternation (`||' or `or').

Negation has highest precedence. Alternation and concatenation have equal precedence and associate left to right. Note that explicit and tokens, not juxtaposition, are now required for concatenation.

If an identifier is given without a keyword, the most recent keyword is assumed. For example,

not host vs and ace

is short for

not host vs and host ace

which should not be confused with

not ( host vs or ace )

Expression arguments can be passed to tcpdump as either a single argument or as multiple arguments, whichever is more convenient. Generally, if the expression contains Shell metacharacters, it is easier to pass it as a single, quoted argument. Multiple arguments are concatenated with spaces before being parsed.

EXAMPLES

To print all packets arriving at or departing from sundown:

tcpdump host sundown

To print traffic between helios and either hot or ace:

tcpdump host helios and \( hot or ace \)

To print all IP packets between ace and any host except helios:

tcpdump ip host ace and not helios

To print all traffic between local hosts and hosts at Berkeley:

tcpdump net ucb-ether

To print all ftp traffic through internet gateway snup: (note that the expression is quoted to prevent the shell from (mis-)interpreting the parentheses):

tcpdump 'gateway snup and (port ftp or ftp-data)'

To print traffic neither sourced from nor destined for local hosts (if you gateway to one other net, this stuff should never make it onto your local net).

tcpdump ip and not net localnet

To print the start and end packets (the SYN and FIN packets) of each TCP conversation that involves a non-local host.

tcpdump 'tcp[tcpflags] & (tcp-syn|tcp-fin) != 0 and not src and dst net localnet'

To print IP packets longer than 576 bytes sent through gateway snup:

tcpdump 'gateway snup and ip[2:2] > 576'

To print IP broadcast or multicast packets that were not sent via ethernet broadcast or multicast:

tcpdump 'ether[0] & 1 = 0 and ip[16] >= 224'

To print all ICMP packets that are not echo requests/replies (i.e., not ping packets):

tcpdump 'icmp[icmptype] != icmp-echo and icmp[icmptype] != icmp-echoreply'

OUTPUT FORMAT

The output of tcpdump is protocol dependent. The following gives a brief description and examples of most of the formats.

Link Level Headers

If the '-e' option is given, the link level header is printed out. On ethernets, the source and destination addresses, protocol, and packet length are printed.

On FDDI networks, the '-e' option causes tcpdump to print the `frame control' field, the source and destination addresses, and the packet length. (The `frame control' field governs the interpretation of the rest of the packet. Normal packets (such as those containing IP datagrams) are `async' packets, with a priority value between 0 and 7; for example, `async4'. Such packets are assumed to contain an 802.2 Logical Link Control (LLC) packet; the LLC header is printed if it is not an ISO datagram or a so-called SNAP packet.

On Token Ring networks, the '-e' option causes tcpdump to print the `access control' and `frame control' fields, the source and destination addresses, and the packet length. As on FDDI networks, packets are assumed to contain an LLC packet. Regardless of whether the '-e' option is specified or not, the source routing information is printed for source-routed packets.

On 802.11 networks, the '-e' option causes tcpdump to print the `frame control' fields, all of the addresses in the 802.11 header, and the packet length. As on FDDI networks, packets are assumed to contain an LLC packet.

(N.B.: The following description assumes familiarity with the SLIP compression algorithm described in RFC-1144.)

On SLIP links, a direction indicator (``I'' for inbound, ``O'' for outbound), packet type, and compression information are printed out. The packet type is printed first. The three types are ip, utcp, and ctcp. No further link information is printed for ip packets. For TCP packets, the connection identifier is printed following the type. If the packet is compressed, its encoded header is printed out. The special cases are printed out as *S+n and *SA+n, where n is the amount by which the sequence number (or sequence number and ack) has changed. If it is not a special case, zero or more changes are printed. A change is indicated by U (urgent pointer), W (window), A (ack), S (sequence number), and I (packet ID), followed by a delta (+n or -n), or a new value (=n). Finally, the amount of data in the packet and compressed header length are printed.

For example, the following line shows an outbound compressed TCP packet, with an implicit connection identifier; the ack has changed by 6, the sequence number by 49, and the packet ID by 6; there are 3 bytes of data and 6 bytes of compressed header:

O ctcp * A+6 S+49 I+6 3 (6)

ARP/RARP Packets

Arp/rarp output shows the type of request and its arguments. The format is intended to be self explanatory. Here is a short sample taken from the start of an `rlogin' from host rtsg to host csam:

arp who-has csam tell rtsg
arp reply csam is-at CSAM

The first line says that rtsg sent an arp packet asking for the ethernet address of internet host csam. Csam replies with its ethernet address (in this example, ethernet addresses are in caps and internet addresses in lower case).

This would look less redundant if we had done tcpdump -n:

arp who-has 128.3.254.6 tell 128.3.254.68
arp reply 128.3.254.6 is-at 02:07:01:00:01:c4

If we had done tcpdump -e, the fact that the first packet is broadcast and the second is point-to-point would be visible:

RTSG Broadcast 0806  64: arp who-has csam tell rtsg
CSAM RTSG 0806  64: arp reply csam is-at CSAM

For the first packet this says the ethernet source address is RTSG, the destination is the ethernet broadcast address, the type field contained hex 0806 (type ETHER_ARP) and the total length was 64 bytes.

TCP Packets

(N.B.:The following description assumes familiarity with the TCP protocol described in RFC-793. If you are not familiar with the protocol, neither this description nor tcpdump will be of much use to you.)

The general format of a tcp protocol line is:

src > dst: flags data-seqno ack window urgent options

Src and dst are the source and destination IP addresses and ports. Flags are some combination of S (SYN), F (FIN), P (PUSH), R (RST), W (ECN CWR) or E (ECN-Echo), or a single `.' (no flags). Data-seqno describes the portion of sequence space covered by the data in this packet (see example below). Ack is sequence number of the next data expected the other direction on this connection. Window is the number of bytes of receive buffer space available the other direction on this connection. Urg indicates there is `urgent' data in the packet. Options are tcp options enclosed in angle brackets (e.g., <mss 1024>).

Src, dst and flags are always present. The other fields depend on the contents of the packet's tcp protocol header and are output only if appropriate.

Here is the opening portion of an rlogin from host rtsg to host csam.

rtsg.1023 > csam.login: S 768512:768512(0) win 4096 <mss 1024>
csam.login > rtsg.1023: S 947648:947648(0) ack 768513 win 4096 <mss 1024>
rtsg.1023 > csam.login: . ack 1 win 4096
rtsg.1023 > csam.login: P 1:2(1) ack 1 win 4096
csam.login > rtsg.1023: . ack 2 win 4096
rtsg.1023 > csam.login: P 2:21(19) ack 1 win 4096
csam.login > rtsg.1023: P 1:2(1) ack 21 win 4077
csam.login > rtsg.1023: P 2:3(1) ack 21 win 4077 urg 1
csam.login > rtsg.1023: P 3:4(1) ack 21 win 4077 urg 1

The first line says that tcp port 1023 on rtsg sent a packet to port login on csam. The S indicates that the SYN flag was set. The packet sequence number was 768512 and it contained no data. (The notation is `first:last(nbytes)' which means `sequence numbers first up to but not including last which is nbytes bytes of user data'.) There was no piggy-backed ack, the available receive window was 4096 bytes and there was a max-segment-size option requesting an mss of 1024 bytes.

Csam replies with a similar packet except it includes a piggy-backed ack for rtsg's SYN. Rtsg then acks csam's SYN. The `.' means no flags were set. The packet contained no data so there is no data sequence number. Note that the ack sequence number is a small integer (1). The first time tcpdump sees a tcp `conversation', it prints the sequence number from the packet. On subsequent packets of the conversation, the difference between the current packet's sequence number and this initial sequence number is printed. This means that sequence numbers after the first can be interpreted as relative byte positions in the conversation's data stream (with the first data byte each direction being `1'). `-S' will override this feature, causing the original sequence numbers to be output.

On the 6th line, rtsg sends csam 19 bytes of data (bytes 2 through 20 in the rtsg -> csam side of the conversation). The PUSH flag is set in the packet. On the 7th line, csam says it's received data sent by rtsg up to but not including byte 21. Most of this data is apparently sitting in the socket buffer since csam's receive window has gotten 19 bytes smaller. Csam also sends one byte of data to rtsg in this packet. On the 8th and 9th lines, csam sends two bytes of urgent, pushed data to rtsg.

If the snapshot was small enough that tcpdump didn't capture the full TCP header, it interprets as much of the header as it can and then reports ``[|tcp]'' to indicate the remainder could not be interpreted. If the header contains a bogus option (one with a length that's either too small or beyond the end of the header), tcpdump reports it as ``[bad opt]'' and does not interpret any further options (since it's impossible to tell where they start). If the header length indicates options are present but the IP datagram length is not long enough for the options to actually be there, tcpdump reports it as ``[bad hdr length]''.

Capturing TCP packets with particular flag combinations (SYN-ACK, URG-ACK, etc.)

There are 8 bits in the control bits section of the TCP header:

CWR | ECE | URG | ACK | PSH | RST | SYN | FIN

Let's assume that we want to watch packets used in establishing a TCP connection. Recall that TCP uses a 3-way handshake protocol when it initializes a new connection; the connection sequence with regard to the TCP control bits is

1) Caller sends SYN

2) Recipient responds with SYN, ACK

3) Caller sends ACK

Now we're interested in capturing packets that have only the SYN bit set (Step 1). Note that we don't want packets from step 2 (SYN-ACK), just a plain initial SYN. What we need is a correct filter expression for tcpdump.

Recall the structure of a TCP header without options:

 0                            15                              31
-----------------------------------------------------------------
|          source port          |       destination port        |
-----------------------------------------------------------------
|                        sequence number                        |
-----------------------------------------------------------------
|                     acknowledgment number                     |
-----------------------------------------------------------------
|  HL   | rsvd  |C|E|U|A|P|R|S|F|        window size            |
-----------------------------------------------------------------
|         TCP checksum          |       urgent pointer          |
-----------------------------------------------------------------

A TCP header usually holds 20 octets of data, unless options are present. The first line of the graph contains octets 0 - 3, the second line shows octets 4 - 7 etc.

Starting to count with 0, the relevant TCP control bits are contained in octet 13:

 0             7|             15|             23|             31
----------------|---------------|---------------|----------------
|  HL   | rsvd  |C|E|U|A|P|R|S|F|        window size            |
----------------|---------------|---------------|----------------
|               |  13th octet   |               |               |

Let's have a closer look at octet no. 13:

                |               |
                |---------------|
                |C|E|U|A|P|R|S|F|
                |---------------|
                |7   5   3     0|

These are the TCP control bits we are interested in. We have numbered the bits in this octet from 0 to 7, right to left, so the PSH bit is bit number 3, while the URG bit is number 5.

Recall that we want to capture packets with only SYN set. Let's see what happens to octet 13 if a TCP datagram arrives with the SYN bit set in its header:

                |C|E|U|A|P|R|S|F|
                |---------------|
                |0 0 0 0 0 0 1 0|
                |---------------|
                |7 6 5 4 3 2 1 0|

Looking at the control bits section we see that only bit number 1 (SYN) is set.

Assuming that octet number 13 is an 8-bit unsigned integer in network byte order, the binary value of this octet is

00000010

and its decimal representation is

   7     6     5     4     3     2     1     0
0*2 + 0*2 + 0*2 + 0*2 + 0*2 + 0*2 + 1*2 + 0*2  =  2

We're almost done, because now we know that if only SYN is set, the value of the 13th octet in the TCP header, when interpreted as a 8-bit unsigned integer in network byte order, must be exactly 2.

This relationship can be expressed as

tcp[13] == 2

We can use this expression as the filter for tcpdump in order to watch packets which have only SYN set:

tcpdump -i xl0 tcp[13] == 2

The expression says "let the 13th octet of a TCP datagram have the decimal value 2", which is exactly what we want.

Now, let's assume that we need to capture SYN packets, but we don't care if ACK or any other TCP control bit is set at the same time. Let's see what happens to octet 13 when a TCP datagram with SYN-ACK set arrives:

     |C|E|U|A|P|R|S|F|
     |---------------|
     |0 0 0 1 0 0 1 0|
     |---------------|
     |7 6 5 4 3 2 1 0|

Now bits 1 and 4 are set in the 13th octet. The binary value of octet 13 is

     00010010

which translates to decimal

   7     6     5     4     3     2     1     0
0*2 + 0*2 + 0*2 + 1*2 + 0*2 + 0*2 + 1*2 + 0*2   = 18

Now we can't just use 'tcp[13] == 18' in the tcpdump filter expression, because that would select only those packets that have SYN-ACK set, but not those with only SYN set. Remember that we don't care if ACK or any other control bit is set as long as SYN is set.

In order to achieve our goal, we need to logically AND the binary value of octet 13 with some other value to preserve the SYN bit. We know that we want SYN to be set in any case, so we'll logically AND the value in the 13th octet with the binary value of a SYN:

          00010010 SYN-ACK              00000010 SYN
     AND  00000010 (we want SYN)   AND  00000010 (we want SYN)
          --------                      --------
     =    00000010                 =    00000010

We see that this AND operation delivers the same result regardless whether ACK or another TCP control bit is set. The decimal representation of the AND value as well as the result of this operation is 2 (binary 00000010), so we know that for packets with SYN set the following relation must hold true:

( ( value of octet 13 ) AND ( 2 ) ) == ( 2 )

This points us to the tcpdump filter expression

     tcpdump -i xl0 'tcp[13] & 2 == 2'

Note that you should use single quotes or a backslash in the expression to hide the AND ('&') special character from the shell.

UDP Packets

UDP format is illustrated by this rwho packet:

actinide.who > broadcast.who: udp 84

This says that port who on host actinide sent a udp datagram to port who on host broadcast, the Internet broadcast address. The packet contained 84 bytes of user data.

Some UDP services are recognized (from the source or destination port number) and the higher level protocol information printed. In particular, Domain Name service requests (RFC-1034/1035) and Sun RPC calls (RFC-1050) to NFS.

UDP Name Server Requests

(N.B.:The following description assumes familiarity with the Domain Service protocol described in RFC-1035. If you are not familiar with the protocol, the following description will appear to be written in greek.)

Name server requests are formatted as

src > dst: id op? flags qtype qclass name (len)

h2opolo.1538 > helios.domain: 3+ A? ucbvax.berkeley.edu. (37)

Host h2opolo asked the domain server on helios for an address record (qtype=A) associated with the name ucbvax.berkeley.edu. The query id was `3'. The `+' indicates the recursion desired flag was set. The query length was 37 bytes, not including the UDP and IP protocol headers. The query operation was the normal one, Query, so the op field was omitted. If the op had been anything else, it would have been printed between the `3' and the `+'. Similarly, the qclass was the normal one, C_IN, and omitted. Any other qclass would have been printed immediately after the `A'.

A few anomalies are checked and may result in extra fields enclosed in square brackets: If a query contains an answer, authority records or additional records section, ancount, nscount, or arcount are printed as `[na]', `[nn]' or `[nau]' where n is the appropriate count. If any of the response bits are set (AA, RA or rcode) or any of the `must be zero' bits are set in bytes two and three, `[b2&3=x]' is printed, where x is the hex value of header bytes two and three.

UDP Name Server Responses

Name server responses are formatted as

src > dst:  id op rcode flags a/n/au type class data (len)

helios.domain > h2opolo.1538: 3 3/3/7 A 128.32.137.3 (273)
helios.domain > h2opolo.1537: 2 NXDomain* 0/1/0 (97)

In the first example, helios responds to query id 3 from h2opolo with 3 answer records, 3 name server records and 7 additional records. The first answer record is type A (address) and its data is internet address 128.32.137.3. The total size of the response was 273 bytes, excluding UDP and IP headers. The op (Query) and response code (NoError) were omitted, as was the class (C_IN) of the A record.

In the second example, helios responds to query 2 with a response code of non-existent domain (NXDomain) with no answers, one name server and no authority records. The `*' indicates that the authoritative answer bit was set. Since there were no answers, no type, class or data were printed.

Other flag characters that might appear are `-' (recursion available, RA, not set) and `|' (truncated message, TC, set). If the `question' section doesn't contain exactly one entry, `[nq]' is printed.

Note that name server requests and responses tend to be large and the default snaplen of 68 bytes may not capture enough of the packet to print. Use the -s flag to increase the snaplen if you need to seriously investigate name server traffic. `-s 128' has worked well for me.

SMB/CIFS decoding

tcpdump now includes fairly extensive SMB/CIFS/NBT decoding for data on UDP/137, UDP/138 and TCP/139. Some primitive decoding of IPX and NetBEUI SMB data is also done.

By default a fairly minimal decode is done, with a much more detailed decode done if -v is used. Be warned that with -v a single SMB packet may take up a page or more, so only use -v if you really want all the gory details.

If you are decoding SMB sessions containing unicode strings then you may wish to set the environment variable USE_UNICODE to 1. A patch to auto-detect unicode strings would be welcome.

For information on SMB packet formats and what all te fields mean see www.cifs.org or the pub/samba/specs/ directory on your favorite samba.org mirror site. The SMB patches were written by Andrew Tridgell (tridge@samba.org).

NFS Requests and Replies

Sun NFS (Network File System) requests and replies are printed as:

src.xid > dst.nfs: len op args
src.nfs > dst.xid: reply stat len op results


sushi.6709 > wrl.nfs: 112 readlink fh 21,24/10.73165
wrl.nfs > sushi.6709: reply ok 40 readlink "../var"
sushi.201b > wrl.nfs:
        144 lookup fh 9,74/4096.6878 "xcolors"
wrl.nfs > sushi.201b:
        reply ok 128 lookup fh 9,74/4134.3150

In the first line, host sushi sends a transaction with id 6709 to wrl (note that the number following the src host is a transaction id, not the source port). The request was 112 bytes, excluding the UDP and IP headers. The operation was a readlink (read symbolic link) on file handle (fh) 21,24/10.731657119. (If one is lucky, as in this case, the file handle can be interpreted as a major,minor device number pair, followed by the inode number and generation number.) Wrl replies `ok' with the contents of the link.

In the third line, sushi asks wrl to lookup the name `xcolors' in directory file 9,74/4096.6878. Note that the data printed depends on the operation type. The format is intended to be self explanatory if read in conjunction with an NFS protocol spec.

If the -v (verbose) flag is given, additional information is printed. For example:

sushi.1372a > wrl.nfs:
        148 read fh 21,11/12.195 8192 bytes @ 24576
wrl.nfs > sushi.1372a:
        reply ok 1472 read REG 100664 ids 417/0 sz 29388

(-v also prints the IP header TTL, ID, length, and fragmentation fields, which have been omitted from this example.) In the first line, sushi asks wrl to read 8192 bytes from file 21,11/12.195, at byte offset 24576. Wrl replies `ok'; the packet shown on the second line is the first fragment of the reply, and hence is only 1472 bytes long (the other bytes will follow in subsequent fragments, but these fragments do not have NFS or even UDP headers and so might not be printed, depending on the filter expression used). Because the -v flag is given, some of the file attributes (which are returned in addition to the file data) are printed: the file type (``REG'', for regular file), the file mode (in octal), the uid and gid, and the file size.

If the -v flag is given more than once, even more details are printed.

Note that NFS requests are very large and much of the detail won't be printed unless snaplen is increased. Try using `-s 192' to watch NFS traffic.

NFS reply packets do not explicitly identify the RPC operation. Instead, tcpdump keeps track of ``recent'' requests, and matches them to the replies using the transaction ID. If a reply does not closely follow the corresponding request, it might not be parsable.

AFS Requests and Replies

Transarc AFS (Andrew File System) requests and replies are printed as:

src.sport > dst.dport: rx packet-type
src.sport > dst.dport: rx packet-type service call call-name args
src.sport > dst.dport: rx packet-type service reply call-name args


elvis.7001 > pike.afsfs:
        rx data fs call rename old fid 536876964/1/1 ".newsrc.new"
        new fid 536876964/1/1 ".newsrc"
pike.afsfs > elvis.7001: rx data fs reply rename

In the first line, host elvis sends a RX packet to pike. This was a RX data packet to the fs (fileserver) service, and is the start of an RPC call. The RPC call was a rename, with the old directory file id of 536876964/1/1 and an old filename of `.newsrc.new', and a new directory file id of 536876964/1/1 and a new filename of `.newsrc'. The host pike responds with a RPC reply to the rename call (which was successful, because it was a data packet and not an abort packet).

In general, all AFS RPCs are decoded at least by RPC call name. Most AFS RPCs have at least some of the arguments decoded (generally only the `interesting' arguments, for some definition of interesting).

The format is intended to be self-describing, but it will probably not be useful to people who are not familiar with the workings of AFS and RX.

If the -v (verbose) flag is given twice, acknowledgement packets and additional header information is printed, such as the the RX call ID, call number, sequence number, serial number, and the RX packet flags.

If the -v flag is given twice, additional information is printed, such as the the RX call ID, serial number, and the RX packet flags. The MTU negotiation information is also printed from RX ack packets.

If the -v flag is given three times, the security index and service id are printed.

Error codes are printed for abort packets, with the exception of Ubik beacon packets (because abort packets are used to signify a yes vote for the Ubik protocol).

Note that AFS requests are very large and many of the arguments won't be printed unless snaplen is increased. Try using `-s 256' to watch AFS traffic.

AFS reply packets do not explicitly identify the RPC operation. Instead, tcpdump keeps track of ``recent'' requests, and matches them to the replies using the call number and service ID. If a reply does not closely follow the corresponding request, it might not be parsable.

KIP AppleTalk (DDP in UDP)

AppleTalk DDP packets encapsulated in UDP datagrams are de-encapsulated and dumped as DDP packets (i.e., all the UDP header information is discarded). The file /etc/atalk.names is used to translate AppleTalk net and node numbers to names. Lines in this file have the form

number  name

1.254           ether
16.1            icsd-net
1.254.110       ace

The first two lines give the names of AppleTalk networks. The third line gives the name of a particular host (a host is distinguished from a net by the 3rd octet in the number - a net number must have two octets and a host number must have three octets.) The number and name should be separated by whitespace (blanks or tabs). The /etc/atalk.names file may contain blank lines or comment lines (lines starting with a `#').

AppleTalk addresses are printed in the form

net.host.port

144.1.209.2 > icsd-net.112.220
office.2 > icsd-net.112.220
jssmag.149.235 > icsd-net.2

(If the /etc/atalk.names doesn't exist or doesn't contain an entry for some AppleTalk host/net number, addresses are printed in numeric form.) In the first example, NBP (DDP port 2) on net 144.1 node 209 is sending to whatever is listening on port 220 of net icsd node 112. The second line is the same except the full name of the source node is known (`office'). The third line is a send from port 235 on net jssmag node 149 to broadcast on the icsd-net NBP port (note that the broadcast address (255) is indicated by a net name with no host number - for this reason it's a good idea to keep node names and net names distinct in /etc/atalk.names).

NBP (name binding protocol) and ATP (AppleTalk transaction protocol) packets have their contents interpreted. Other protocols just dump the protocol name (or number if no name is registered for the protocol) and packet size.

NBP packets are formatted like the following examples:

icsd-net.112.220 > jssmag.2: nbp-lkup 190: "=:LaserWriter@*"
jssmag.209.2 > icsd-net.112.220: nbp-reply 190: "RM1140:LaserWriter@*" 250
techpit.2 > icsd-net.112.220: nbp-reply 190: "techpit:LaserWriter@*" 186

The first line is a name lookup request for laserwriters sent by net icsd host 112 and broadcast on net jssmag. The nbp id for the lookup is 190. The second line shows a reply for this request (note that it has the same id) from host jssmag.209 saying that it has a laserwriter resource named "RM1140" registered on port 250. The third line is another reply to the same request saying host techpit has laserwriter "techpit" registered on port 186.

ATP packet formatting is demonstrated by the following example:

jssmag.209.165 > helios.132: atp-req  12266<0-7> 0xae030001
helios.132 > jssmag.209.165: atp-resp 12266:0 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:1 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:2 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:4 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:6 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp*12266:7 (512) 0xae040000
jssmag.209.165 > helios.132: atp-req  12266<3,5> 0xae030001
helios.132 > jssmag.209.165: atp-resp 12266:3 (512) 0xae040000
helios.132 > jssmag.209.165: atp-resp 12266:5 (512) 0xae040000
jssmag.209.165 > helios.132: atp-rel  12266<0-7> 0xae030001
jssmag.209.133 > helios.132: atp-req* 12267<0-7> 0xae030002

Jssmag.209 initiates transaction id 12266 with host helios by requesting up to 8 packets (the `<0-7>'). The hex number at the end of the line is the value of the `userdata' field in the request.

Helios responds with 8 512-byte packets. The `:digit' following the transaction id gives the packet sequence number in the transaction and the number in parens is the amount of data in the packet, excluding the atp header. The `*' on packet 7 indicates that the EOM bit was set.

Jssmag.209 then requests that packets 3 & 5 be retransmitted. Helios resends them then jssmag.209 releases the transaction. Finally, jssmag.209 initiates the next request. The `*' on the request indicates that XO (`exactly once') was not set.

IP Fragmentation

Fragmented Internet datagrams are printed as

(frag id:size@offset+)
(frag id:size@offset)

(The first form indicates there are more fragments. The second indicates this is the last fragment.)

Id is the fragment id. Size is the fragment size (in bytes) excluding the IP header. Offset is this fragment's offset (in bytes) in the original datagram.

The fragment information is output for each fragment. The first fragment contains the higher level protocol header and the frag info is printed after the protocol info. Fragments after the first contain no higher level protocol header and the frag info is printed after the source and destination addresses. For example, here is part of an ftp from arizona.edu to lbl-rtsg.arpa over a CSNET connection that doesn't appear to handle 576 byte datagrams:

arizona.ftp-data > rtsg.1170: . 1024:1332(308) ack 1 win 4096 (frag 595a:328@0+)
arizona > rtsg: (frag 595a:204@328)
rtsg.1170 > arizona.ftp-data: . ack 1536 win 2560

There are a couple of things to note here: First, addresses in the 2nd line don't include port numbers. This is because the TCP protocol information is all in the first fragment and we have no idea what the port or sequence numbers are when we print the later fragments. Second, the tcp sequence information in the first line is printed as if there were 308 bytes of user data when, in fact, there are 512 bytes (308 in the first frag and 204 in the second). If you are looking for holes in the sequence space or trying to match up acks with packets, this can fool you.

A packet with the IP don't fragment flag is marked with a trailing (DF).

Timestamps

By default, all output lines are preceded by a timestamp. The timestamp is the current clock time in the form

hh:mm:ss.frac

and is as accurate as the kernel's clock. The timestamp reflects the time the kernel first saw the packet. No attempt is made to account for the time lag between when the ethernet interface removed the packet from the wire and when the kernel serviced the `new packet' interrupt.  

SEE ALSO

stty(1), pcap(3), bpf(4), nit(4P), pfconfig(8)