Resume downloading removed from queue nzb

[jimage]

I've been using the beta3 and was about 1gb the way into downloading a file when the RSS that I was messing with decided to add 10 pages of downloads to the queue (I'm hoping beta4 fixes this because no matter how I've set up the rss it eventually decides to send all the backlog of the rss to the queue) so I purged all rather than meticulously removing 10 pages of crap.  Well now my 1gb is stranded and I'm not sure how to resume it, adding the same nzb just starts from scratch with a new incomplete temp folder.  Is there a way of resuming?

[shypike]

In such cases it's safer to just clear the queue.
Startup with the --clean option. The History will no longer be erased by this.

I don't understand your backlog problem.
The initial batch of a new RSS feed is marked (*) so that it's not downloaded.
All subsequent scans will download matching new NZBs.
If you haven't updates an existing feed for a while, the backlog may be large.

[jimage]

I don't understand either, when I first load the rss I check the preview and see that there are matched files with * that supposedly won't be downloaded.  So I enable the rss and nothing is immediately added.  But then when I think everything is fine I look down at some point maybe a few hours later and it's trying to download the whole rss.

Any ideas?  And any idea how I can resume my download that got cut off?

[shypike]

I'll setup a test for the RSS.

Resuming your download is not easy.
Is it still listed in the queue?
If not, you should probably download it again.
The alternative is to find the part that was downloaded in the temporary download folder,
download the missing files seperately, move them in one folder
and run par2 and unrar manually.

[jimage]

Thanks for your help, it's a shame resuming has to be like that, I was hoping it wouldn't come to that.

Oh and the rss was from nzbindex.nl if it helps.

[shypike]

SABnzbd is about automating downloading.
This is only possible if the average quality of the NZB files is good.
That's my we concentrate on automation features
and compromised on the interactive features.
SABnzbd is not about manually collecting scraping together stuff from unreliable sources.
There are better programs for that, which are in turn much less automation friendly.

Accidents like yours are unfortunate, but part of the risk of using Beta's.
Sorry about that.

[jimage]

Just to follow up a bit, I installed the new beta4 and the same error is occurring.  This time I left my house for a while and came back and not only had sabnzbd downloaded the files with a * from the rss but had downloaded each file four times and probably would have kept going.  Makes absolutely no sense.

[shypike]

Since Beta4 was already in pre-release and your complaint is the only one,
we had no time to look at this.
Are you sure nzbindex.nl doesn't suddenly add more content?
Can you email your specific RSS URL to bugs at sabnzbd.org ?

[jimage]

Definitely not adding new content.  I check the preview in sabnzbd before I enable the rss and then after and it includes the exact same files.  I sent an email with everything, screenshots etc.  Please get back to me if you need anything else.

[shypike]

The explanation of the problem is actually quite easy.
nzbindex.nl generates new URLs for every RSS readout.

In the example below, the boldface part changes all the time.

http://www.nzbindex.nl/download/15122899-1261088031/bladibla.nzb

So obviously, SABnzbd thinks there are new jobs and keeps happily adding them to the queue.
The simplest work-around is to use the option "Prevent Duplicate Downloads" in combination
with an NZB backup folder.

This problem is solvable in SABnzbd, but only by including extra code specific for nzbindex.nl.
This is not a very desirable solution, because where do we end?

I'll discuss this with the team.

[jimage]

Ah okay that makes sense, thanks for the help, I can use another rss for now I'm sure, or maybe I'll setup the no duplicates.

Sorry to keep asking questions but back on my original question a little, so sabnzbd loses the queue when shut down? (computer restart or manual sabnzbd shutdown)  Is this supposed to happen or should I just stop playing with the beta and switch to the stable?  I thought only me purging the queue would make me have to re-download unfinished nzbs, but shutting down sabnzbd or restarted has the same effect.

[shypike]

The queue is preserved when a proper shutdown is done.
Manual shutdown (through the menu) is always OK.

Letting the OS shutdown SABnzbd is potentially problematic.
When SABnzbd is actually downloading and has a full memory cache,
the shutdown takes time because so much must be written to disk.
If the OS isn't patient enough to wait for SABnzbd's shutdown
but resorts to a blunt process termination, chances are that the queue
admin will be corrupt.

[jimage]

Ah okay, well a Windows 7 Pro restart seems to force sabnzbd to close too soon because I always loose my queue that way, although the manual shutdown is working just fine now (I had thought I was having a problem with it but I was wrong).

And I had no idea you could click the link of the nzb in the queue and show/edit the individual files.  I should have guessed but now I'm glad that I know.  I can reinsert the nzb that I mentioned that got lost after purging the queue and only download the files I need.  I don't mind piecing it together later with quickpar and winrar, but wouldn't it be nice to have an option for sabnzbd to put files in a same folder rather than adding the .# to it if there is a folder there already? A merge folder kinda thing. Or am I missing another feature?

[shypike]

No, merging is not an option, because you might have jobs with the same name.
There will be a major redesign in the next half year, that will allow better recovery.

Moving files from one folder to another manually, isn't a big deal I think.