You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
$ sudo -u maa-worker ./maa-run.sh update
Fetching MaaCore version info (channel: stable)...
Error: Failed to load MaaCore!
Caused by:
libMaaCore.so: cannot open shared object file: No such file or directory
$ sudo -u maa-worker ./maa-run.sh install
Fetching MaaCore version info (channel: stable)...
Downloading MaaCore 5.6.0...
Already downloaded, skip downloading
Installing MaaCore...
Extracting archive file...
Done!
Already up to date.
由于我用来运行
maa-cli
的云服务器的硬盘空间有限,我把maa-cli
和maa-core
都放到了一个nfs
文件系统中,今天检查服务器日志发现服务器在定时执行maa update
时发生了报错再次尝试执行
maa update
也无法进行(提示MaaCore
无法加载),只能执行maa install
,应该是第一次执行失败时remove_dir_all
实际上已经清空了library
文件夹导致的在网上也能搜索到少量类似反馈,有
rust
的,也有python
库shutil
的,我怀疑是linux中的nfs客户端有缺陷,在递归删除文件夹中的文件时有延迟导致会概率发生这个问题类似的问题在我这里已经不是第一次出现了,还好这次我及时发现了,不然晚点基建又得崩😭
根据这个问题的表现,感觉
maa-cli
在实现ensure_clean
方法时进行一次重试就能规避这个bug,不知道是否方便The text was updated successfully, but these errors were encountered: