[Bug] Fail to import Web Bnowledge Base but get SUCCESS (Importing the WEB knowledge base failed, but it showed success)
Contact Information
MaxKB Version
社区版 v1.8.0 (build at 2024-12-05T12:08, commit: a3e7fcb5)
Problem Description
Recently, I wanted to deploy DeepSeek+maxkb+ollama, but encountered this error when configuring the knowledge base. I didn't realize there was an error URL in my document before, so it stopped after importing 11 items. But I got feedback SUCCESS, and I hope to modify it. At the same time, I hope to add the funIction of automatically skipping error URLs, which should provide great convenience.
Steps to Reproduce
- Install maxkb according to the maxkb manual. I don't think this is necessary, but I still mention it.
- Load the ollama model according to the maxkb manual, the same as above.
- Build a knowledge base. Since I don't have too many documents locally, I used the document summary on the website. Here is the address
- I directly processed it with Notepad, and I have attached the processed file(20 rows).
Shows success, same as edge devtools
But obviously there are many documents missing.
Then I found that this address is invalid.
I have also ruled out the reason related to the selector,
Finally, I hope to add a function to automatically skip errors. It would be best if the error URL can be returned.
Attach: proceed knowledge base.txt
- 根据maxkb手册安装maxkb, 我不认为这有必要, 但还是提及.
- 根据maxkb手册加载ollama模型, 和上面同样.
- 构建知识库, 由于我本地没有太多的文档,使用了网站上的文档汇总, 这里是地址 4.我直接用notepad处理了, 另附20条txt文档.如下
显示成功, edge devtools同样
但明显缺少很多文档.
随后发现这个地址(暂时只看了一条)失效了.
我也排除了选择器的原因,
附件: proceed knowledge base.txt
The expected correct result
Fix this feedback bug.
Finally, I hope to add a function to automatically skip errors. It would be best if the error URL can be returned.
修复返回bug
最后, 希望增加一个自动跳过错误的功能, 如果可以返回错误URL,那是最好了.
Related log output
Additional Information
No response
您好,感谢反馈。导入文档如果遇到地址返回错误,会被忽略不会提示错误信息,不过这里确实存在一个问题,就是碰到错误地址后,后续导入会被停止,导致导入不完整。我们会在后续版本修复这个问题。
Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑🤝🧑👫🧑🏿🤝🧑🏻👩🏾🤝👨🏿👬🏿
Hello, thanks for the feedback. If an error is encountered when importing a document, it will be ignored and will not prompt an error message. However, there is indeed a problem here, that is, after encountering the wrong address, the subsequent import will be stopped, resulting in incomplete import. We will fix this issue in subsequent versions.
V1.10.3-lts版本已发布,请更新试用!
Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑🤝🧑👫🧑🏿🤝🧑🏻👩🏾🤝👨🏿👬🏿
V1.10.3-lts version has been released, please update and try it out!