Queue hangs at 100% (sabnzbd 0.7.4)

[djmartain]

Since a week, whenever i startup my nas (QNAP TS-439 II+ with 1GB ram and x86 atom cpu) and sickbeard starts adding my shows to SABNZBD, something goes wrong. The problem is that, some of these downloads just hang at 100% and other downloads don't download at all, they just disappear and i have to find them in the queue repair screen. When i add them again to the queue the download immediately fails saying "out of server retention because 0% can be downloaded".

With both situations i cannot get the downloads to complete UNLESS i remove the corrupt queue items (the 100% stuck I cannot remove but it is gone after a restart) and re-add and restart the downloads.

I have already tried repairing the queue but without success.


One final unrelated bug I encounter is that the tv sorting check-box cannot stay checked. Every time sab restarts, that check-box is back unchecked. As long as sab keeps running, everything is fine. All the other settings get updated and saved correctly after restart.

Any suggestions?

edit: found a few interessting lines in the logs:

Code: Select all

2012-10-16 12:49:58,210::INFO::[nzbqueue:204] Saving queue
2012-10-16 12:49:58,227::INFO::[__init__:896] Saving data for queue9.sab in /share/MD0_DATA/.qpkg/SABnzbdplus/Config/admin/queue9.sab
2012-10-16 12:49:58,229::INFO::[nzbqueue:535] Sorting by average date...(reversed:False)
2012-10-16 12:49:58,230::ERROR::[nzbqueue:197] Failed to restart NZB after pre-check (SABnzbd_nzo_HnvZtK)
2012-10-16 12:49:58,231::INFO::[nzbqueue:198] Traceback: 
Traceback (most recent call last):
  File "/share/MD0_DATA/.qpkg/SABnzbdplus/sabnzbd/nzbqueue.py", line 193, in replace_in_q
    self.__nzo_list[targetpos] = new_nzo
IndexError: list assignment index out of range

[shypike]

You are using pre-check?
The log segment is definitely interesting, because obviously that indicates a bug.
Do the failing jobs eventually turn out not to be on the server? Or is this just an result of other problems?
Can you email one of the troublemaking NZB files to [email protected] ?
(Assuming it's about specific ones).

I cannot confirm the TV Sort checkbox problem.
Maybe you should reset your browser cache. And then refresh the page fully.
One thing you can check is whether the sabnzbd.ini file is updated after you check the box and save.
(Make sure you change something or the file will not be updated).
Look for the line
tv_sort = 1

[djmartain]

I am using pre-check indeed. The failing jobs are just released nzb's and are brand new and complete on server. I don't think it is about a specific nzb but i will send one i am having problems with.

[kiljaden]

I am using pre-check indeed. The failing jobs are just released nzb's and are brand new and complete on server. I don't think it is about a specific nzb but i will send one i am having problems with.

I've been experiencing the same issues like you. So you could is a know issue that shypike is looking at. A temporary work around is to disable the pre-check function (it works for me at least).

[shypike]

I think I have uncovered the issue when it happens just after a pre-check.
The fix will be in 0.7.5

[djmartain]

Great news, looking forward to this version. I'll probably disable pre-checking until it's out.

[miles267]

Thanks guys for identifying this issue. I too was experiencing the same issue and just came on to research it.

[thomasmaerz]

When will 0.7.4 be released? I'm also having the 100% job stuck issue:

tried to repair the queue, got this:

Code: Select all

2012-10-27 10:40:30,151	ERROR:	Loading /home/niggaplz/.sabnzbd/admin/future/SABnzbd_attrib failed

Here's my log file for the relevent time up until I tried to do the queue rebuild:

Code: Select all

2012-10-27 10:40:29,599::DEBUG::[__init__:821] Saving data for SABnzbd_nzo_wlSvEW in /home/niggaplz/Downloads/incomplete/Black Swan 2010 BluRay 1080p DTS x264 CHD.1/__ADMIN__
2012-10-27 10:40:29,603::INFO::[__init__:896] Saving data for queue9.sab in /home/niggaplz/.sabnzbd/admin/queue9.sab
2012-10-27 10:40:29,604::DEBUG::[notifier:112] Sending registration to localhost:23053
2012-10-27 10:40:29,604::DEBUG::[notifier:219] To : localhost:23053 <<class 'gntp.GNTPRegister'>>
2012-10-27 10:40:29,605::DEBUG::[growler:171] Cannot register with Growl [Errno 111] Connection refused
2012-10-27 10:40:30,117::DEBUG::[__init__:855] Loading data for SABnzbd_attrib from /home/niggaplz/.sabnzbd/admin/future/SABnzbd_attrib
2012-10-27 10:40:30,151::ERROR::[__init__:871] Loading /home/niggaplz/.sabnzbd/admin/future/SABnzbd_attrib failed
2012-10-27 10:40:30,151::INFO::[__init__:872] Traceback: 
Traceback (most recent call last):
  File "/usr/share/sabnzbdplus/sabnzbd/__init__.py", line 863, in load_data
    data = cPickle.load(_f)
UnpicklingError: invalid load key, '*'.
2012-10-27 10:40:30,165::DEBUG::[downloader:124] Initializing downloader/decoder
2012-10-27 10:40:30,166::INFO::[__init__:920] Loading data for watched_data.sab from /home/niggaplz/.sabnzbd/admin/watched_data.sab
2012-10-27 10:40:30,168::DEBUG::[scheduler:138] Scheduling RSS interval task every 60 min (delay=38)
2012-10-27 10:40:30,168::INFO::[__init__:322] All processes started
2012-10-27 10:40:30,169::INFO::[sabnzbdplus:334] Web dir is /usr/share/sabnzbdplus/interfaces/Plush
2012-10-27 10:40:30,169::INFO::[sabnzbdplus:334] Web dir is /usr/share/sabnzbdplus/interfaces/Config
2012-10-27 10:40:30,197::INFO::[sabnzbdplus:450] _yenc module... found!
2012-10-27 10:40:30,197::INFO::[sabnzbdplus:458] par2 binary... found (/usr/bin/par2)
2012-10-27 10:40:30,197::INFO::[sabnzbdplus:466] unrar binary... found (/usr/bin/unrar)
2012-10-27 10:40:30,197::INFO::[sabnzbdplus:471] unzip binary... found (/usr/bin/unzip)
2012-10-27 10:40:30,198::INFO::[sabnzbdplus:477] nice binary... found (/usr/bin/nice)
2012-10-27 10:40:30,198::INFO::[sabnzbdplus:481] ionice binary... found (/usr/bin/ionice)
2012-10-27 10:40:30,198::INFO::[sabnzbdplus:486] pyOpenSSL... found (True)
2012-10-27 10:40:30,199::INFO::[sabnzbdplus:1473] Starting web-interface on 0.0.0.0:9999
2012-10-27 10:40:30,200::INFO::[_cplogging:55] [27/Oct/2012:10:40:30] ENGINE Bus STARTING
2012-10-27 10:40:30,203::INFO::[_cplogging:55] [27/Oct/2012:10:40:30] ENGINE Started monitor thread '_TimeoutMonitor'.
2012-10-27 10:40:30,417::INFO::[_cplogging:55] [27/Oct/2012:10:40:30] ENGINE Serving on 0.0.0.0:9999
2012-10-27 10:40:30,418::INFO::[_cplogging:55] [27/Oct/2012:10:40:30] ENGINE Bus STARTED
2012-10-27 10:40:30,420::DEBUG::[notifier:112] Sending registration to localhost:23053
2012-10-27 10:40:30,421::DEBUG::[notifier:219] To : localhost:23053 <<class 'gntp.GNTPRegister'>>
2012-10-27 10:40:30,421::DEBUG::[growler:171] Cannot register with Growl [Errno 111] Connection refused
2012-10-27 10:40:30,920::DEBUG::[misc:553] Version check disabled by Ubuntu/Debian packaging
2012-10-27 10:40:30,921::INFO::[sabnzbdplus:1542] Starting sabnzbdplus-0.7.4
2012-10-27 10:40:30,921::DEBUG::[__init__:333] Starting postprocessor
2012-10-27 10:40:30,922::DEBUG::[__init__:336] Starting assembler
2012-10-27 10:40:30,923::DEBUG::[__init__:339] Starting downloader
2012-10-27 10:40:30,924::DEBUG::[scheduler:181] Starting scheduler
2012-10-27 10:40:30,926::DEBUG::[__init__:344] Starting dirscanner
2012-10-27 10:40:30,926::INFO::[dirscanner:265] Dirscanner starting up
2012-10-27 10:40:30,927::DEBUG::[__init__:349] Starting urlgrabber
2012-10-27 10:40:30,928::INFO::[urlgrabber:77] URLGrabber starting up

[shypike]

Disable "check before download" until 0.7.5 comes out.

[thomasmaerz]

What does "check before download" do exactly?

It was never enabled on my install. Tried turning it on and my stuff started downloading.

Is there an ETA on 0.7.5 release?

[shypike]

If the option wasn't enabled, you're not hit by its bug.
It checks presence of all articles on the server, before actually downloading.
Most people don't need that.
Can you answer the same question posted here?
https://forums.sabnzbd.org/viewtopic.php?f=3&t=12076

[rao0oar]

I am having the same issue. My check before downloading is turned off, and has been for some time. also my download speed is crazy slow, but the rest of my computer the speed is fine.??

[thomasmaerz]

Is one job blocking the queue?
Have you looked at the job that should finish?
When you click on the job, you'll get the list of remaining files. Are there any?
What is the "only get articles from the top of the queue" setting? It should be off.

One job is not blocking the queue right now, but it has before. Right now my stalled jobs have files left to download, but I've had some previously with no files left to download. Those seem to have finally finished after I moved them farther down the queue and they sat for awhile. "only get articles from the top of the queue" is and has been set to off.

[shypike]

I am already examining jobs with zero files still going and I'm close to a solution.
When you have jobs hanging which still have files left, it's more likely that it has to do with stalling servers.
Can you see in the Status page if there are still connections with servers?

[thomasmaerz]

I am already examining jobs with zero files still going and I'm close to a solution.
When you have jobs hanging which still have files left, it's more likely that it has to do with stalling servers.
Can you see in the Status page if there are still connections with servers?

i will check that out this evening and post the result.