queue hangs on download.php

[escii]

Hey there,

My current setup is sickbeard and sabnzbd running on the synology ds209, I used the following setup guide http://kevin.vanzonneveld.net/techblog/ ... _synology/.

Ever since the installation files seem to hang in the sabnzbd queue with something similar to the below image.



I have checked the logs and sickbeard says the below. Seems as if sickbeard grabs the episode and then passees to sab no problems though the download sits in the queue endlessly like the image above until I manually delete the link and fetch the files a second time.

Sickbeard log

Code: Select all

Mar-14 15:07:13 INFO     SEARCHQUEUE-RSS-SEARCH :: Clearing NZBMatrix cache and updating with new information
Mar-14 15:07:10 INFO     SEARCHQUEUE-RSS-SEARCH :: Searching all providers for any needed episodes
Mar-14 15:07:10 INFO     SEARCHQUEUE-RSS-SEARCH :: Beginning search for new episodes on RSS
Mar-14 15:07:10 INFO     SEARCHQUEUE-RSS-SEARCH :: Changing all old missing episodes to status WANTED
Mar-14 14:07:16 INFO     SEARCHQUEUE-RSS-SEARCH :: Sending NZB to SABnzbd
Mar-14 14:07:16 INFO     SEARCHQUEUE-RSS-SEARCH :: Quality of File Name is SD TV
Mar-14 14:07:16 INFO     SEARCHQUEUE-RSS-SEARCH :: Found result File Name at https://api.nzbmatrix.com/v1.1/download.php?id=1211156&username=&apikey=&scenename=1
Mar-14 14:07:13 INFO     SEARCHQUEUE-RSS-SEARCH :: Clearing NZBMatrix cache and updating with new information
Mar-14 14:07:10 INFO     SEARCHQUEUE-RSS-SEARCH :: Searching all providers for any needed episodes
Mar-14 14:07:10 INFO     SEARCHQUEUE-RSS-SEARCH :: Beginning search for new episodes on RSS

SABnzbd log

Code: Select all

2012-03-14 13:55:32,767::INFO::[rss:376] Running scheduled RSS scan
2012-03-14 14:07:16,610::INFO::[__init__:380] Fetching https://api.nzbmatrix.com/v1.1/download.php?id=1211156&username=&apikey=&scenename=1
2012-03-14 14:07:16,615::INFO::[nzbqueue:92] Saving queue
2012-03-14 14:07:16,679::INFO::[urlgrabber:126] Grabbing URL https://api.nzbmatrix.com/v1.1/download.php?id=1211156&username=&apikey
2012-03-14 14:07:49,232::INFO::[nzbqueue:92] Saving queue
2012-03-14 14:07:49,269::INFO::[__init__:702] /root/.sabnzbd/cache/SABnzbd_nzo__PZDL3 removed
2012-03-14 14:07:49,272::INFO::[nzbqueue:92] Saving queue
2012-03-14 14:55:32,774::INFO::[rss:376] Running scheduled RSS scan
2012-03-14 15:55:32,787::INFO::[rss:376] Running scheduled RSS scan
2012-03-14 16:38:31,788::INFO::[__init__:702] /root/.sabnzbd/cache/SABnzbd_nzo_bUtmRk removed
2012-03-14 16:38:31,790::INFO::[nzbqueue:92] Saving queue
2012-03-14 16:39:14,505::INFO::[__init__:380] Fetching https://api.nzbmatrix.com/v1.1/download.php?id=1211156&username=&apikey=&scenename=1
2012-03-14 16:39:14,510::INFO::[nzbqueue:92] Saving queue
2012-03-14 16:39:14,577::INFO::[urlgrabber:126] Grabbing URL https://api.nzbmatrix.com/v1.1/download.php?id=1211156&username=&apikey=

Anyone know a fix to this particular issue?

[shypike]

I assume that you edited SABnzbd's log?
If not, you're missing the nzbmatrix apikey.
Other than that, nzbmatrix is not always reliable and sometimes
SABnzbd get completely confused by some error situations.
We're still looking at this.

[escii]

Correct I made an edit to the log. I had previously been running the windows version of SAB and sick without errors such as the above.

Is this something to do with the linux version running as a daemon?

[shypike]

Is this something to do with the linux version running as a daemon?

Don't think so, occasionally Windows users have the complaint.
It seems to be a combination of nzbmatrix malfunctioning and
SABnzbd not handling it properly somehow.
Still under investigation.

[escii]

Whats the difference in the application when the auto RSS feed finds the NZB compared to a manual fetch? Seems it should be the same behaviour...

[shypike]

It is the same behaviour.
However sometime an entry that has a failure will have trouble finishing.
The frequency of rss entries is much higher than your manual actions.

[escii]

So there is know workaraound and nothing which can be tweaked, basically waiting until a fix is found and I should forget about the automation possibilities within sabnzbd?

[shypike]

The problem is simply that these errors occur not very frequently and definitely not on demand.
So far I have never been able to reproduce these scenarios so I cannot find out what exactly goes wrong.
What I have been able to gather so far is that it's always triggered by errors on the indexer's side.
SABnzbd should recover from that (at least when the error is not persistent)
but sometimes doesn't do this very well.
Another complication is that SABnzbd cannot really see the difference between a transient
and a permanent error. For example when nzbmatrix delivers an empty NZB or just an HTML page,
SABnzbd cannot know whether retrying makes sense or not. Retrying forever is not a solution.

The only exception is newzbin where I recently discovered a badly handled error situation (fix will be in 0.7.0).

[shypike]

Release 0.7.0 will have a different approach to retries.

- Unlimited retries (until you delete the item from the queue)
- Failed items will have a waiting time depending on the type of error
(e.g. time-out: 60 seconds, partial NZB: 120 sec, no nzb: 600 sec)
- Failed items will be sent to the back of the queue (not visibly in the download queue, but internally)

[escii]

Funnily enough sab is working correctly since sunday... although I have seen two nzb's fault in the way I ve described in this thread but now something different which I think might be related to sick, the nzb quality is set to SD in sick but comes down the pipe as x264. although unrelated to this topic, any thoughts on this? also when will the new release be available.

[shypike]

Soon, a limited group has started testing.

[escii]

Cool, the replies are appreciated. Thanks.

[MatchkeY]

Am a Windows user with exact same problem.
Seemed to start happening after I updated to the latest release (Sorry not sure what release I was on before, but was running for ~ 12 months Id say).

Testing a few things, clean launch etc. and will get back to this thread if I resolve it.
Else I'm looking forward to 0.7.0

[TonioRoffo]

Same issue here.

XBMClive (installed) with sabnzbd+ still at 0.5.0, getting nzb's from sickbeard. Occasional problems with "resolving" the download.php!id links coming from nzbmatrix. I know nzbmatrix has a throttle on the time between lookups. In the sickbeard log you can see some lines referring to thottling to respect the nzbmatrix rules.

On the website I can find this error to be possible:

error:please_wait_x = Please wait x seconds before retry.

This is probably what causes it?

[escii]

I wonder if this problem has something to do with the following post on the sick forum http://sickbeard.com/forums/viewtopic.php?f=3&t=3917 and the possibility sab's getting confused with the nzb sicks passing.

This thread on sick also answers my query about the ep quality suddenly changing due to the cappers file format, im finding some epz are coming down in the x264 rather than SD quality... anyway back on topic still wonder if this causing the queue to hang?