Agent Vinod created NET-460:
-------------------------------

             Summary: _retrieveFile() blocks calling thread, on FTP I/O till 
the time file transfer is complete
                 Key: NET-460
                 URL: https://issues.apache.org/jira/browse/NET-460
             Project: Commons Net
          Issue Type: Wish
          Components: FTP
    Affects Versions: 3.1
         Environment: linux/windows
            Reporter: Agent Vinod


The Function _retrieveFile in file: FTPClient.java , does not respond to 
interrupts from calling thread.

For Example:

A Basic FTP Client Application has 1 Main (Parent) Thread and 1 Child Thread.
Main (Parent) thread handles all functions except the FtpClient download/upload.
Child Thread handles only FtpClient related functions mainly (_retrieveFile()) 
etc.

Steps to reproduce:

1) Main Thread has initiated child Thread .
2) Child thread is presently downloading a file using _retrieveFile(String 
command, String remote, OutputStream local) . 
3) After some time, Main Thread fires Interrupt on child Thread to stop( Abort) 
download.

Expected behavior:

Child Thread immediately aborts download and dies.

Observed behavior:

Child Thread blocks on retrieveFile(String command, String remote, OutputStream 
local) till the file finishes download. 
Only after this ,does the Child thread respond to any interrupt from Parent 
Thread.

My Workaround:

file:  FTPClient.java
Class: FTPClient    

Step 1: declare private Socket mySocket;

Step 2: In the function : protected boolean _retrieveFile(String command, 
String remote, OutputStream local) throws IOException{}
Comment out:  Socket socket;
and instead use:  mySocket ( declared as global in step1)

Step 3:  In the function : public boolean abort() throws IOException
Add a statement: Util.closeQuietly(mySocket);
before the statement: return FTPReply.isPositiveCompletion(abor());


This way, every time the Main Thread calls abort(), the download active and 
blocked on mySocket in _retrieveFile() is immediately interrupted and stopped.
raising an immediate Exception and thus stopping the Child thread (of course 
one needs to catch this exception properly).

I am not sure if this is the right way of doing it and am afraid if this breaks 
something else.
Requesting the core developers to look into a better solution to this 
workaround.

thank




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to