opendataio / hcfsfuse Goto Github PK

Describe
In the scenario of reading a 300MB file from cos,fnifuse takes 30s, but fnrfuse takes 12s

Fix #13050

As the README.md said, after
mvn clean package
command, hcfsfuse--jar-with-dependencies.jar was generated in the folder target.
However, when I try
$ java -jar target/hcfsfuse-1.0.0-SNAPSHOT-jar-with-dependencies.jar -c core-site.xml -c another-site.xml -m /Users/mbl/fusefs -r https://fuse-optimize-xxx.cos.ap-chengdu.myqcloud.com
, in which the url is the address of COS bucket. There occurs ERRORS:
[main] WARN util.NativeCodeLoader (NativeCodeLoader.java:<clinit>) - Unable to load native-hadoop library for your platform... using builtin-java classes where applicable [main] INFO fuse.HCFSFuse (HCFSFuse.java:main) - mounting to /Users/mbl/fusefs [Thread-6] ERROR fuse.AlluxioFuseUtils (AlluxioFuseUtils.java:getIdInfo) - Failed to get id from with option -u [Thread-6] ERROR fuse.AlluxioFuseUtils (AlluxioFuseUtils.java:getGidFromGroupName) - Failed to get gid from group name .
while the process is NOT interrupted, the shell is stuck and finally the connection to server shutdown.
How can I solve this problem? Is something wrong with the URL of my COS or the access authority need to be changed?

I have a S3 bucket and I want to use hcfsfuse to mount a local directory to the bucket.
But I can't find the way to this thing. Please give me some help.
Thank you

Thanks for your great work!
I have few questions on it.

1. I think there is no need to import spring related dependencies.

2. Why not using maven shade plugin. Because when i use hcfsuse to mount hdfs, found that i can't get the hdfs filesystem through Java SPI. related question: https://stackoverflow.com/questions/17265002/hadoop-no-filesystem-for-scheme-file. When i refactor the way of packaging, it works well.

jar /tmp/hcfsfuse-1.0.0-SNAPSHOT.jar -m /tmp/m3 -r hdfs://XXXX
ls /tmp

and i got:

d????????? ? ? ? ? ? m3