JDownloader Community - Appwork GmbH
 

Reply
 
Thread Tools Display Modes
  #1  
Old 24.12.2009, 05:58
drbits's Avatar
drbits drbits is offline
JD English Support (inactive)
 
Join Date: Sep 2009
Location: Physically in Los Angeles, CA, USA
Posts: 4,437
Default Server does not support chunkload (9.579)

This relates to http://board.jdownloader.org/showthread.php?t=12072

In this case, the log shows a download that returned after only 15KB of a 200MB file. I have no evidence, but this could well be an HTML error page (the suspected problem is that the download quota for the hour was exceeded). Because this is at the beginning of the file, we could examine the file signature to determine whether this is part of the file or something else.

In at least two cases (in Nightly), I have downloaded a RAR file beginning with an HTML error page, followed by the Rar data (missing the end of the original file).

When a chunk read returns less than 50KB (less than 10 seconds on dialup), should JD retain the data or should it start the next download for the file at the same position?
---------------------------
If this was an HTML page, shouldn't JD have identified that? Evaluating the HTML might be harder and this is not a common host. However some simple regular expressions might work for almost all hosts. For example:
.*[wW][aA][iI][tT]\w?([0-9]+)\w?([hH]ours?^O[mM]inutes?^O[sS]econds?).*
%1 and %2 should identify the length of time to wait (should JD add 10%?).
(Please forgive me if this is wrong, it has been almost 15 years since I made heavy use of *nix regex).
---------------------
In this case, JD tried to continue reading from the last byte it had downloaded. The response was other than a chunkload header and that information is in the log, however the actual header is not there (and would be useful, it could have identified the actual error).

The result was the "Server does not support chunkload" error status. However, this was ignored and less than 10 seconds later, the same chunk download (with the same offset) was attempted (with the same results).
---------------------
Would it be hard to add a command line option that would log the first 16 (or 64) bytes of each new file and all HTTP headers with a status other than 200?

I realize that the JD status subsystem is complex and this might not be easy. However, actually evaluating headers and confirming file headers could lead to a significant quality improvement.

My personal opinion is that if JD evaluated this information when an error occurs, much better status information could be reported and separating Permanent Errors, Temporary Errors, and Retryable Errors would be possible. In addition, it would make it easier to determine whether an error is specific to a link, a host, or a problem with JD or the connection.
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

All times are GMT +2. The time now is 15:17.
Provided By AppWork GmbH | Privacy | Imprint
Parts of the Design are used from Kirsch designed by Andrew & Austin
Powered by vBulletin® Version 3.8.10 Beta 1
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.