In your package ， Conncetion Class is not public ， how can I use it outside the package ？

When I find a file ，I must use two params ( volum id and magic number) as fid .
How I want to use myself uniqu_id (just like uuid) ，and I also want to use this uuid that it can find out file
Thx ！

帮忙回答一下 多谢

FileSource fileSource = new FileSource(); // SeaweedFS master server host fileSource.setHost("192.168.100.114"); // SeaweedFS master server portaaaaa fileSource.setPort(9335); // Startup manager and listens for the change try { fileSource.startup(); } catch (IOException e) { e.printStackTrace(); } FileTemplate template = new FileTemplate(fileSource.getConnection()); File file = new File("G:\\007.JPG"); FileHandleStatus status = template.saveFileByStream("007.jpg", new FileInputStream(file));

that's my code get this error
org.ehcache.spi.serialization.UnsupportedTypeException: No serializer found for type LookupVolumeResult.class
so i rewrite LookupVolumeResult to implements java.io.Serializable the problem sloved

This is the code that I have right now, and it puzzles me. It doesn't work. I'm sure I have SeaweedFS set up correctly because I got pyseaweed to work perfectly: https://github.com/utek/pyseaweed.

import org.lokra.seaweedfs.core.FileSource;
import org.lokra.seaweedfs.core.FileTemplate;

import java.io.*;

public class Hello {
    public static void main(String[] args) {
        System.out.println("Hello, World");

        FileSource fileSource = new FileSource();
        fileSource.setHost("localhost");
        fileSource.setPort(9333);
        try {
            fileSource.startup();
            System.out.println("Finished starting up!");
        } catch (IOException e) {
            System.out.println("Aiya! Cannot connect!");
            return;
        }

        try {
            InputStream is = new FileInputStream("/Users/Pac/Code/golang/flow.jpg");
            FileTemplate template = new FileTemplate(fileSource.getConnection());
            template.saveFileByStream("flow.jpg", is);
        } catch(FileNotFoundException e) {
            System.out.println("Cannot save the file!");
            return;
        } catch(IOException e) {
            System.out.println("Cannot write to Seaweed!");
            return;
        }

        System.out.println("Success and done!");
    }
}

This is the console output.

Hello, World
Nov 23, 2016 11:00:11 PM org.lokra.seaweedfs.core.FileSource startup
INFO: start connect to the seaweedfs core server [http://localhost:9333]
Nov 23, 2016 11:00:13 PM org.lokra.seaweedfs.core.Connection startup
INFO: core connection is startup now
Finished starting up!
329 [main] INFO org.ehcache.core.EhcacheManager - Cache 'lookupVolumeCache' created in EhcacheManager.
Cannot write to Seaweed!

Hello,

I'm interested in using your library in a production system, and was wondering if there's a developer guide somewhere I couldn't find? Your library looks promising.

Whenever you request a file and try to get the FileHandleStatus and after that the FileStream to i.e. compose a response with both combined, there will be a NPE thrown at org.lokra.seaweedfs.core.Connection.fetchStreamCacheByRequest(Connection.java:328).

public class SeaweedFile {
	@JsonProperty
	public FileHandleStatus status;
	@JsonProperty
	public String data;
}

@GET
@Path("{id}")
public Response read(@PathParam("id") String fileID) {
	Response response = null;
	FileTemplate template = new FileTemplate(fileSource.getConnection());
	try {
		SeaweedFile file = new SeaweedFile();
		file.setStatus(template.getFileStatus(fileID)); // first get the FileStatus  *X
		
		StreamResponse fileStream = template.getFileStream(fileID); // *Y
		InputStream is = fileStream.getInputStream();
		file.setData(new String(ByteStreams.toByteArray(is)));
               // do return stuff ...
       } ...
}

If you switch lines marked with *X and *Y, the request will return smoothly.

i have multiple master how to use works

seaweedfs use http head method to get file information.

Headers:

Accept-Ranges →bytes
Content-Disposition →inline; filename="1.txt"
Content-Encoding →gzip
Content-Length →686
Content-Type →text/plain; charset=utf-8
Date →Mon, 03 Jul 2017 06:45:16 GMT
Etag →"1e1b55d9"
Last-Modified →Mon, 03 Jul 2017 03:05:04 GMT

seaweedfs-client code here

FileTemplate#getFileStatus(String)
headerResponse.getLastHeader("Content-Disposition").getValue().substring(10, headerResponse.getLastHeader("Content-Disposition").getValue().length() - 1)

resolved result is : lename="1.txt

maybe it can be

Header contentDisposition = headerResponse.getLastHeader("Content-Disposition");
HeaderElement[] elements = contentDisposition.getElements();
if(ArrayUtils.isEmpty(elements))  { // org.apache.commons.lang3.ArrayUtils
  return "";
}else {
  return elements[elements.length-1].getValue();
}

weedfs on docker, " /dir/assign" return "url / publicUrl" are inner IP, how to modify volume url?

line 267 of Connection.java:
EntityUtils.consume(entity);

this line of code will close the InputStream of reponse entity instantly, which cause the InputStream returned throw Exception when read.
So you simply cannot read a file if reponse body is larger than fileStreamCacheSize!

It won't work if you just delete this line of code. The connection is closed in the finally clause. So the InputStream returned by FileTemplate.getFileStream will always be closed when you try to read it.

The fix is to determine whether the reponse is cacheable(smaller than fileStreamCacheSize), if is not, read the response body into memory and return a InputStream point to the memory.

Does this tool support upload large file by using weed large file handling steps?
https://github.com/chrislusf/seaweedfs/wiki/Large-File-Handling
I cannot find the method for uploading json string.
thx~

In Connection.java:, there are 4 places found with code request.setHeader("Connection", "close");

This line of code disabled the keep-alive feature, so the connection will be discarded instead of pooled.

I don't know why did the author explicitly disabled keep-alive feature while using a PooledHttpClient.