jd-opensource / jd-opensource/joyagent-jdgenie
部署问题集合参考
- Dominant language
- Java
- Stars
- 11.9k
- Forks
- 1.7k
- PR merge metrics
- No merged PRs in 30d
Description
# 1. pnpm install -> zsh: command not found: pnpm
如何安装pnpm,https://pnpm.io/zh/installation
ps:资源包的安装如果速度比较慢建议使用国内镜像
# 2. The operation couldn’t be completed. Unable to locate a Java Runtime.
mac用户安装
*. brew install maven
直接安装会同时安装openjava sdk, JAVA_HOME:/opt/homebrew/Cellar/openjdk/24.0.1/libexec/openjdk.jdk/Contents/Home, 如果没有需要手动安装java.
**.下载 https://www.oracle.com/java/technologies/downloads/,版本大于17
手动安装:/usr/libexec/java_home -V
/Library/Java/JavaVirtualMachines/jdk-17.jdk/Contents/Home
然后通过,写入zshrc
echo 'export JAVA_HOME=$(/usr/libexec/java_home)' >> ~/.zshrc
echo 'export PATH=$JAVA_HOME/bin:$PATH' >> ~/.zshrc
source ~/.zshrc
# 3. 请求拒绝或者失败
查看文件https://github.com/jd-opensource/joyagent-jdgenie/blob/main/genie-backend/src/main/resources/application.yml
一般失败的原因LLM请求失败,以及后端服务端口不一致。
1. https://github.com/jd-opensource/joyagent-jdgenie/blob/main/genie-backend/src/main/resources/application.yml#L13 其中对于LLM请求标准的openai 格式:https://api.openai.com/v1
其中对于claude, 我们测试发现claude3.7模型在react模式下反思和工具调用,效果要好于openai一些模型,所以为了达到更好的效果建议使用claude模型在react模式。配置建议如下:
settings: '{"claude-3-7-sonnet-v1": {
"model": "claude-3-7-sonnet-20250219",
"max_tokens": 8192,
"temperature": 0,
"base_url": "",
"apikey": "",
"max_input_tokens": 128000
}}'
2. https://github.com/jd-opensource/joyagent-jdgenie/blob/main/genie-backend/src/main/resources/application.yml#L96 可以在这里配置后端工具以及mcp服务端口(如果不一致的话)
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by reviewing the troubleshooting topics in this issue and the referenced genie-backend/src/main/resources/application.yml entries around lines 13 and 96. Verify the pnpm, Java, LLM, and backend/MCP port guidance against the linked resources. Done means the setup problems and their remedies are organized in a project documentation location, although this issue does not name that location.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- java
- Domain
- devops, documentation
- Issue type
- Documentation
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100