blocks|key|717877|text|尝试使用tcpdump，定义一个只监听您的程序正在通信的端口/地址的过滤器，并汇总打印的长度属性。|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|entityRanges|data|717878|entityMap^0|4|7|0^^$0|@$1|2|3|4|5|6|7|H|8|@$9|I|A|J|B|C]]|D|@]|E|$]]|$1|F|3|-4|5|6|7|K|8|@]|D|@]|E|$]]]|G|$]]

Try <code>tcpdump</code>, define a filter that will listen only on the port/address that your program is communicating on and sum up the printed length attribute.

blocks|key|717906|text|使用Kernel+Probes，您可以动态地“闯入”内核/系统调用，并读取/记录它们的参数。对read()/recv*()和/或write()/send*()执行此操作可能会对您有所帮助。|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|entityRanges|data|717907|entityMap|0|LINK|mutability|MUTABLE|url|https://www.kernel.org/doc/Documentation/kprobes.txt^0|1B|6|1I|7|1S|7|20|7|2|D|0|0^^$0|@$1|2|3|4|5|6|7|N|8|@$9|O|A|P|B|C]|$9|Q|A|R|B|C]|$9|S|A|T|B|C]|$9|U|A|V|B|C]]|D|@$9|W|A|X|1|Y]]|E|$]]|$1|F|3|-4|5|6|7|Z|8|@]|D|@]|E|$]]]|G|$H|$5|I|J|K|E|$L|M]]]]

Using <a href="https://www.kernel.org/doc/Documentation/kprobes.txt" rel="nofollow">Kernel Probes</a> you can "break into" kernel/system calls on the fly and read/log their arguments. Doing this for <code>read()</code>/<code>recv*()</code> and/or <code>write()</code>/<code>send*()</code> might help you.

blocks|key|2637743|text|您可能可以查看/proc/net/tcp文件的内容来读取一些Linux+/send统计数据，在大多数系统上，+netstat实用程序执行相同的操作，但方式更简洁。|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|BOLD|entityRanges|data|2637744|在您的程序中，在不同的时间点解析这个文件，以收集在套接字上接收/发送的字节的信息，尽管我不确定它是否可以将字节映射到特定的TCP套接字。|2637745|entityMap^0|7|D|1J|7|1E|4|0|0^^$0|@$1|2|3|4|5|6|7|K|8|@$9|L|A|M|B|C]|$9|N|A|O|B|C]|$9|P|A|Q|B|D]]|E|@]|F|$]]|$1|G|3|H|5|6|7|R|8|@]|E|@]|F|$]]|$1|I|3|-4|5|6|7|S|8|@]|E|@]|F|$]]]|J|$]]

You could probably look at the contents of <code>/proc/net/tcp</code> file to read some of the recv/send statistics <code>netstat</code> utility on most Linux systems does the same thing but in a cleaner way.

In you program, parse this file at various points of time to gather information on bytes received/sent on a socket, though i am not sure if it can map the bytes for a specific TCP socket.

blocks|key|4029378|text|不确定您是否正在查看套接字，或者是否需要计算给定接口上的RX/TX字节的总量。我需要解决后一个问题，我提供了可能对您有用的以下脚本：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|4029379|while+true;+do+MARKER=$(ifconfig+eth0+%7C+grep+"[RT]X+bytes"+%7C+awk+'{print+$2}'+%7C+cut+-c+7-);+sleep+10;+NEWMARKER=$(ifconfig+eth0+%7C+grep+"[RT]X+bytes"+%7C+awk+'{print+$2}'+%7C+cut+-c+7-);+expr+${NEWMARKER}+-+${MARKER};+set+$MARKER=$NEWMARKER;+done|code-block|syntax|javascript|4029380|请注意，对我来说，接收的字节数实际上是唯一重要的值，如果需要，可以将其用作模板来提取发送的字节。|4029381|将休眠值更改为所需的任何时间间隔(以秒为单位)，如果需要，请重定向到文件。|4029382|entityMap^0|0|0|0|0^^$0|@$1|2|3|4|5|6|7|M|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|N|8|@]|9|@]|A|$E|F]]|$1|G|3|H|5|6|7|O|8|@]|9|@]|A|$]]|$1|I|3|J|5|6|7|P|8|@]|9|@]|A|$]]|$1|K|3|-4|5|6|7|Q|8|@]|9|@]|A|$]]]|L|$]]

Not certain if you're looking at sockets or need to calculate the whole amount of RX/TX bytes on a given interface. I needed to address the latter problem, whipped up the following script that might be useful to you:

<pre><code>while true; do MARKER=$(ifconfig eth0 | grep "[RT]X bytes" | awk '{print $2}' | cut -c 7-); sleep 10; NEWMARKER=$(ifconfig eth0 | grep "[RT]X bytes" | awk '{print $2}' | cut -c 7-); expr ${NEWMARKER} - ${MARKER}; set $MARKER=$NEWMARKER; done
</code></pre>

Note that for myself the number of received bytes was really the only important value, use as template to extract sent bytes if required.

Change the sleep value to whatever interval you require (measured in seconds), redirect to file if needed.

blocks|key|4029431|text|我要说的是，目前最有趣的方式是使用eBPF在tcp_sendmsg+(用于发送)和tcp_cleanup_rbuf(用于接收)处注入KProbe。|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|entityRanges|data|4029432|有关示例(和工作代码)，请查看BCC+project：https://github.com/iovisor/bcc/blob/master/tools/tcptop.py#L160中的tcptop工具|4029433|从引用的代码中，忽略IPV6如下所示：|4029434|int+kprobe__tcp_sendmsg(struct+pt_regs+*ctx,+struct+sock+*sk,
++++struct+msghdr+*msg,+size_t+size)
{
++++u32+pid+=+bpf_get_current_pid_tgid();
++++u16+dport+=+0,+family+=+sk->__sk_common.skc_family;
++++u64+*val,+zero+=+0;

++++if+(family+!=++AF_INET)+{
++++++return+0;
++++}

++++struct+ipv4_key_t+ipv4_key+=+{.pid+=+pid};

++++ipv4_key.saddr+=+sk->__sk_common.skc_rcv_saddr;
++++ipv4_key.daddr+=+sk->__sk_common.skc_daddr;
++++ipv4_key.lport+=+sk->__sk_common.skc_num;

++++dport+=+sk->__sk_common.skc_dport;
++++ipv4_key.dport+=+ntohs(dport);

++++//+pick+the+entry+in+the+map
++++val+=+ipv4_send_bytes.lookup_or_init(&ipv4_key,+&zero);

++++//+write+to+the+entry+in+the+map
++++(*val)+%2B=+size;
++++return+0;
}|code-block|syntax|javascript|4029435|希望它能帮上忙！|4029436|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/iovisor/bcc|1|https://github.com/iovisor/bcc/blob/master/tools/tcptop.py#L160^0|M|B|15|G|0|2K|6|F|B|0|R|1R|1|0|0|0|0^^$0|@$1|2|3|4|5|6|7|10|8|@$9|11|A|12|B|C]|$9|13|A|14|B|C]]|D|@]|E|$]]|$1|F|3|G|5|6|7|15|8|@$9|16|A|17|B|C]]|D|@$9|18|A|19|1|1A]|$9|1B|A|1C|1|1D]]|E|$]]|$1|H|3|I|5|6|7|1E|8|@]|D|@]|E|$]]|$1|J|3|K|5|L|7|1F|8|@]|D|@]|E|$M|N]]|$1|O|3|P|5|6|7|1G|8|@]|D|@]|E|$]]|$1|Q|3|-4|5|6|7|1H|8|@]|D|@]|E|$]]]|R|$S|$5|T|U|V|E|$W|X]]|Y|$5|T|U|V|E|$W|Z]]]]

I'd say that today the most interesting way of doing this is injecting a KProbe at <code>tcp_sendmsg</code> (for transmit) and <code>tcp_cleanup_rbuf</code>(for receive) using eBPF.

For an example (and working code), check out the <code>tcptop</code> tool from the <a href="https://github.com/iovisor/bcc" rel="nofollow noreferrer">BCC project</a>: <a href="https://github.com/iovisor/bcc/blob/master/tools/tcptop.py#L160" rel="nofollow noreferrer">https://github.com/iovisor/bcc/blob/master/tools/tcptop.py#L160</a> 

From the referenced code, ignoring IPV6 is looks like this:

<pre><code>int kprobe__tcp_sendmsg(struct pt_regs *ctx, struct sock *sk,
 struct msghdr *msg, size_t size)
{
 u32 pid = bpf_get_current_pid_tgid();
 u16 dport = 0, family = sk-&gt;__sk_common.skc_family;
 u64 *val, zero = 0;

 if (family != AF_INET) {
 return 0;
 }

 struct ipv4_key_t ipv4_key = {.pid = pid};

 ipv4_key.saddr = sk-&gt;__sk_common.skc_rcv_saddr;
 ipv4_key.daddr = sk-&gt;__sk_common.skc_daddr;
 ipv4_key.lport = sk-&gt;__sk_common.skc_num;

 dport = sk-&gt;__sk_common.skc_dport;
 ipv4_key.dport = ntohs(dport);

 // pick the entry in the map
 val = ipv4_send_bytes.lookup_or_init(&amp;ipv4_key, &amp;zero);

 // write to the entry in the map
 (*val) += size;
 return 0;
}
</code></pre>

Hope it helps!

For benchmarking I am currently looking for a way to benchmark some TCP-Connections that seem to have problems. For this I would like to log some performance indicators while the program is running. I am already logging the time for each operation, however this just tells me the transmission takes way too long for the use case.

What I would need is a simple way to track the number of bytes read/written at periodic intervals. Normally I would just add the benchmarking manually, but the socket is used in a library so I cannot instrument the read/write calls easily. My guess would be that there is some readily available method provided by linux, that could do this, but I cannot find it.

I already looked at the possibilities of <code>getsockopt()</code> with the <code>TCP_INFO</code> flag, but from the documentation I can only see how to get the window size etc. from this, but not the number of bytes read/written.

Is there any way to get this information from Linux?

How to find out the number of transmitted/received bytes for a socket

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

对于基准测试，我目前正在寻找一种方法来对一些似乎有问题的TCP连接进行基准测试。为此，我想在程序运行时记录一些性能指标。我已经记录了每个操作的时间，但是这只是告诉我传输对于用例来说太长了。我需要的是一种简单的方法来跟踪定期读取/写入的字节数。通常我会手动添加基准测试，但是套接字是在库中使用的，所以我不能轻松地检测读/写...

问如何找出套接字的发送/接收字节数
EN

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问如何找出套接字的发送/接收字节数EN