> For the complete documentation index, see [llms.txt](https://tarscloud.gitbook.io/tarsdocs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://tarscloud.gitbook.io/tarsdocs/tarscpp/tars-performance.md).

# 性能数据

我们测试机器的配置如下图所示，我们共有三台配置相同的机器，其中两台作为客户端，一台作为服务端：

![](/files/rpvCTCtfQXqU1b0VFtEF)

机器配置如下

| 软件     | 软件要求                                    |
| ------ | --------------------------------------- |
| 系统内核   | Linux 3.10.94 x86\_64 x86\_64 GNU/Linux |
| cpu 信息 | 3.30Ghz cpu\*8                          |
| 内存信息   | 16GB                                    |
| 网卡     | 千兆网卡                                    |
| 硬盘信息   | 7200rpm 机械硬盘                            |

第一项测试我们测试机器的常规性能，我们在客户端机器上上部署了2个网络线程和5个业务线程，发送常规大小的业务数据测试 第二项测试我们测试tars的极限性能，我们分别在两台作为客户端的机器上上开了50个进程，每个进程下有16个线程，每个线程发送4字节大小的业务数据包，测试结果如下，

| 语言种类 | 客户端机器数量 | 进程数量 | 线程数量 | 字节包大小（B） | TPS（w/s） | 平均耗时(ms) | cpu利用率 | 网卡入流量(Mb/s) | 出流量(Mb/s) | 网卡入包量(/s) | 出包量(/s) |
| ---- | ------- | ---- | ---- | -------- | -------- | -------- | ------ | ----------- | --------- | --------- | ------- |
| c++  | 1       | 1    | 1    | 4        | 0.7      | 0.13     | 3%     | 8M　　　　　     | 　7M       | 7214　　    | 　7220   |
| c++  | 1       | 1    | 1    | 1024     | 0.52     | 0.17     | 3%     | 53M　　　　     | 　50M      | 5677 　    | 　5666   |
| c++  | 2       | 1    | 10   | 4        | 11.39    | 0.17     | 28%    | 139.729M    | 92.389M   | 127,267   | 127,923 |
| c++  | 2       | 1    | 10   | 1024     | 6.73     | 0.29     | 19%    | 644.135M    | 616.395M  | 95,424    | 95,823  |
| c++  | 2       | 2    | 10   | 4        | 20.05    | 0.20     | 63%    | 238.542M    | 158.658M  | 220,434   | 221,658 |
| c++  | 2       | 2    | 10   | 1024     | 10.03    | 0.38     | 38%    | 972.232M    | 930.256M  | 141,841   | 142,388 |
| c++  | 2       | 5    | 10   | 4        | 27.22    | 0.37     | 84%    | 327.972M    | 215.173M  | 306,896   | 300,099 |
| c++  | 2       | 5    | 10   | 1024     | 10.02    | 0.96     | 48%    | 974.102M    | 932.277M  | 132,091   | 133,574 |
| c++  | 2       | 50   | 16   | 4        | 41       | 3.88     | 92%    | 463.815M    | 313.112M  | 422,732   | 431,050 |
| java | 1       | 1    | 1    | 4        | 0.7      | 0.13     | 5%     | 8.424M　　　　　 | 　6.041M   | 7773　　    | 　7793   |
| java | 1       | 1    | 1    | 1024     | 0.52     | 0.17     | 8%     | 61.866M     | 59.951M   | 6674 　    | 　6700   |
| java | 2       | 1    | 10   | 4        | 9.83     | 0.18     | 34%    | 155.719M    | 106.310M  | 156,681   | 148,201 |
| java | 2       | 1    | 10   | 1024     | 7.82     | 0.27     | 26%    | 694.184M    | 669.369M  | 103,564   | 104,158 |
| java | 2       | 2    | 10   | 4        | 20.42    | 0.20     | 57%    | 254.149M    | 183.307M  | 252,928   | 259,064 |
| java | 2       | 2    | 10   | 1024     | 10.03    | 0.38     | 41%    | 964.790M    | 930.363M  | 141,965   | 143,004 |
| java | 2       | 5    | 10   | 4        | 26.34    | 0.34     | 77%    | 244.887M    | 186.358M  | 243,527   | 254,967 |
| java | 2       | 5    | 10   | 1024     | 10.11    | 0.97     | 48%    | 967.217M    | 939.408M  | 132,421   | 135,919 |
| java | 2       | 50   | 16   | 4        | 38       | 4.27     | 82%    | 438.999M    | 329.996M  | 413,046   | 426,961 |

**测试结果仅供参考，实际结果会因为测试条件和测量方法的不同而产生影响。**


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://tarscloud.gitbook.io/tarsdocs/tarscpp/tars-performance.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.