[功能异常] What is the cause of failed to retrieve data?

[burnacacc700]

问题描述

清晰简洁地描述该错误是什么。

A clear and concise description of what the bug is.

failed to retrieve data on all xhs posts

重现步骤

重现该问题的步骤：

Steps to reproduce the behavior:

... paste link https://www.xiaohongshu.com/explore/6617ee66000000000401ac07?xsec_token=ABMndS0bHg67XtBfan46JYMQGUaBNbz_LW1Jh7vdfXLMo=&xsec_source=pc_user
...rednote awaiting processing
...failed to retrieve data, failed to download remote work files
预期结果

清晰简洁地描述您预期会发生的情况。

A clear and concise description of what you expected to happen.

The video should have downloaded

补充信息

在此添加有关该问题的任何其他上下文信息，例如：操作系统、运行方式、配置文件、错误截图、运行日志等。

请注意：提供配置文件时，请删除 Cookie 内容，避免敏感数据泄露！

Add any other contextual information about the issue here, such as operating system, runtime mode, configuration files,
error screenshots, runtime logs, etc.

Passing cookies does not fix it

I made an issue 1 month ago and was told it was platform limitation, but I waited a month, tried downloading from different accounts, using vpns etc.

Could you help me understand what i need to do? This seems to affect whatever video I try to download, am wondering if my ip/region is blocked?

Thank you.

[dosubot]

The "failed to retrieve data" error is almost always caused by Xiaohongshu's platform restrictions and anti-crawling measures, not by a bug in XHS-Downloader. Xiaohongshu now requires your network environment to be a mainland China IP (not VPN/proxy/overseas), and sometimes even a valid browser Cookie and User-Agent. Even with these, some content—especially older posts or videos—may still be inaccessible due to server-side blocks or recent API changes. Frequent requests can also trigger temporary bans, which usually resolve after about 24 hours. There is currently no universal workaround; some content will remain unavailable until the tool is updated to adapt to new platform changes. More details and troubleshooting steps are discussed here: issue 256, issue 264, and issue 266.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}   

[burnacacc700]

Oh. Does this mean the program is unsuable outside of china, even if logged in?

[dosubot]

Yes, currently the program is not usable from outside mainland China, even if you are logged in. Xiaohongshu enforces strict regional/IP restrictions—your network must be a mainland China ISP IP, or you will encounter CAPTCHA and extraction failures, regardless of cookies or login status. VPNs and proxies are usually detected and blocked, so they don't help bypass this limitation. This is a platform-level restriction, not a bug in the tool, and there are no known workarounds at this time. More details are discussed in issue 215 and issue 248.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}   

[JoeanAmier]

If you can browse the works normally in the browser, you can consider using the Tampermonkey user script.

[JoeanAmier]

I cannot browse the content of the work even after opening the link you provided. It may also be caused by the link to the works.

[burnacacc700]

I did a bit of experimenting and it seems with cookies, it is possible to download, but it gets flagged for too frequent access almost immediately.

Is there another way to slow down the requests? timeout is set at 300

[JoeanAmier]

I did a bit of experimenting and it seems with cookies, it is possible to download, but it gets flagged for too frequent access almost immediately.

Is there another way to slow down the requests? timeout is set at 300

If the link prompts frequent access quickly, the reason for triggering risk control may not be the frequency of access. Increasing the delay may not be effective. For foreign users, will normal access also prompt frequent access?

[burnacacc700]

Normal access meaning access in the web browser?

I think using the program too fast got my account flagged, now in web browser I get the frequent access problem. Yesterday there was no problem.

[JoeanAmier]

You can modify the request delay in source/module/tools.py, which defaults to a random delay of 1 second to 2.5 seconds.

async def sleep_time(
    min_time: int | float = 1.0,
    max_time: int | float = 2.5,
):
    await sleep(uniform(min_time, max_time))

Running with executable files does not support modification, only running through source code or Docker supports modification.

[burnacacc700]

It seems to be working now, but should I be worried about this?

Network error, sssss_2024-06-10_23.38.42
240609Jessica香港fm_应援视频 合唱 西卡。.会一直在这QSy_Jessicatjessica郑秀妍［话题I##郑秀妍香港fml话题J#download failed,ezroz message:RemoteProtocoJEzzoz（peer closed connection without sending complete message body（receive。
32505856 bytes, expected 245492555) ')

[burnacacc700]

I wanted to keep using the exe so instead of sleep I changed the byte size to 512kb