hellanzb uploads as much as it downloads

[Roderik van der Veer]

Binary package hint: hellanzb

I've been using hellanzb for years, as i've been using ubuntu, but the last few weeks i'm running into some strange problem. Downloading a nzb will not run over 15kb/s while downloading from http/ftp runs at 400kb/s

So far this is what i ruled out:

1. the newsserver from my isp would over time sometimes work badly, so i renewed my eweka.nl payed server account, no improvement.
2. I re-installed hardy final yesterday, did apt-get install hellanzb and redid the entire config file. (added server and changed the path's to my download disk, see attachment)
3. Checked my router and disabled QOS. Noticed it was sending as much kb/s as it recieves (what would explain the speed problem since i'm on adsl), why i'm not sure. Verified the router data with iptraf.
4. Installed a newsgroup leech program on my windows laptop, used the same nzb file -> 400+ kb/s (behind the same router, using the same newsserver)
5. installed hardy final on the laptop from step 4, same problem as the desktop. So i really believe there is a problem with hellanzb or on of its deps

so now i'm confused as to what the problem can be, and how i can solve it... The only thing i didn't try was using another newsleech program on hardy, because i don't know any.

[Roderik van der Veer (roderik)]

Forgot to mention the first few alpha and beta releases of hardy worked fine, only in the or one of the last beta's the problems began.

[Jelle De Loecker (skerit)]

Someone posted a fix for this (it was just one line of code missing, something about group updating) but I can't find it anywhere. Anyone, help?

[Caspar Clemens Mierau (leitmedium)]

Thank you for writing this bug report. As I cannot yet reproduce it: Does this error still occur?

[Jelle De Loecker (skerit)]

I reinstalled Ubuntu friday afternoon, 24/10/2008

I only tried a new usenet download through hellanzb yesterday, but I can confirm the bug is still there!

This still fixed it for me:

http://ubuntuforums.org/showpost.php?p=5157098&postcount=3

[Dimitrios Symeonidis (azimout)]

Attaching the resolution as described in ubuntuforums:

 Solution that worked for me.
I ran into what sounds like the same bug after installing Hellanzb on Kubuntu 8.04 (Hardy Heron) (similar to this and this bug report).

The problem was basically that the reported download speed was very low (1/10th the speed it should be), and that Hellanzb was uploading as much as it was downloading. Moreover, the debug log (which can be found at ~/.hellanzb/log-debug if you've activated it) had repeated error messages like:
Code:

2008-06-09 20:47:27,625 giganews[6] getting GROUP: alt.binaries.t
2008-06-09 20:47:27,625 giganews[8] got GROUP: alt.binaries.t
2008-06-09 20:47:27,626 giganews[8] getting GROUP: alt.binaries.t
2008-06-09 20:47:27,627 giganews[2] got GROUP: alt.binaries.t
2008-06-09 20:47:27,627 giganews[2] getting GROUP: alt.binaries.t
2008-06-09 20:47:27,629 giganews[9] got GROUP: alt.binaries.t
2008-06-09 20:47:27,629 giganews[9] getting GROUP: alt.binaries.t
2008-06-09 20:47:27,631 giganews[4] got GROUP: alt.binaries.t
2008-06-09 20:47:27,631 giganews[4] getting GROUP: alt.binaries.t
2008-06-09 20:47:27,636 giganews[3] got GROUP: alt.binaries.t

Basically Hellanzb is repeatedly trying to resolve a newsgroup, but is failing.

The solution (at least for my problem) is described in this bug report. Basically there is a bug in the file:
Code:

/usr/share/python-support/hellanzb/Hellanzb/NZBLeecher/Protocol.py

where it is not cleaning the "groups" string properly. To fix it, edit that file, and below line 513, add a single line of python code that simply does "group = group.strip()". So:
Code:

$ sudo nano /usr/share/python-support/hellanzb/Hellanzb/NZBLeecher/Protocol.py

Then find the code near line 513 and change it like so:
Code:

       if not self.factory.skipGroupCmd:
            # Change group
            for group in self.currentSegment.nzbFile.groups:

                # Added to fix bug
                group = group.strip()

                # NOTE: we could get away with activating only one of the groups instead
                # of all
                if group not in self.activeGroups and group not in self.failedGroups:
                    debug(str(self) + ' getting GROUP: ' + group)
                    self.fetchGroup(group)
                    return

The above fix worked perfectly for me. Hope that helps someone else.