downloads randomly stop, and SABnzbd will not shut down on FreeBSD

[beyonder]

I have a FreeBSD box on which I would like to run SABnzbd+, but I have run into a few issues.  The first issue is that in the FreeBSD ports collection, the version of SABnzbd+ is too old, and has issues with the session key(known issue with new versions of cheetah).  I fixed that by updating to 0.4.12.  Now my downloads will run for a while, but eventually I will get an error that looks like this
2009-10-26 15:23:48,177 WARNING [decoder] CRC Error in (73648759 -> 7F8164FA)

When this happens, a connection will hang, and never work again until I restart SABnzbd+.  Which would be livable, except the fact that when it happens SABnzbd+ will not shut down(I have also tried the 'Force Disconnect' button to no avail).  Clicking shutdown in the Web UI just makes it take up 100% of my CPU.  Sending it a soft kill signal does the same thing(but also prints a warning that a kill signal was caught, and that it will begin shutting down).  Eventually, I send it a nice kill -9, and it goes away.  Start it up again, it downloads for a while... and then it happens again.

Any suggestions?



Version: 0.4.12
OS: FreeBSD 8.0-CURRENT
Install-type: modified ports build(to use 0.4.12 source instead of 0.4.11)
Skin (if applicable): Default
Firewall Software: pf
Are you using IPV6? no
Is the issue reproducible? yes/no. it happens often, but i cannot trigger it.

[shypike]

The warning itself isn't the problem. This is just the standard way of warning about a CRC error in an article.
SABnzbd is dependent on a whole stack of Python modules.
For preferred versions, see: http://sabnzbd.wikidot.com/install-off-modules
We do not directly support FreeBSD, so it's a bit difficult to help.

[beyonder]

Ah, my python is 2.6, and cheetah is 2.2. I will try the recommended versions this weekend(or later this week if I can find the time).  Hopefully, that will fix everything. Thanks!

[shypike]

Cheetah is OK, but we recommend  against using Python 2.6 with SABnzbd 0.4.12.

[ZeroGhost]

I got the same issue as the topic starter does! however there accruing in a shorter and shorter time frame, down to the point where i cant even download 1 episide anymore....
Ive checked both logs at the moment it dies.. nothing in there nothing but normal operational stuff about connections and fileparts, only error in cherry was that he couldnt find a fav icon. so basicly the logs know nothing..
i tried IE to view it in (to rule out firefox causeing something) same problem server stops downloading.
So basically  i got the same issues as the topic poster, unable to shutdown from the webinterface.
the weird part is that this got worse over a week or so.. from like once a day to now: close to un-useble! (i can download for about 10 seconds before it dies).
however the webinterface remains usable.. i can go into all the links and it will work (except the shutdown one.. that starts to initialize it.. and then never stops trying).


(copy pasted from beyonder, but edited where needed)
Version: 0.4.12 (tied 11 aswel, same problem!)
OS: winXP pro x86 SP3 (up to date)
Install-type: Setup installer
Skin (if applicable): tried Default and Smpl both yielded the same result
Firewall Software: Comodo (also tried to turn it off all together, doesn't help.)
Are you using IPV6? yes
Is the issue reproducible? yes/no. it happens often, but i cannot trigger it.


[update]
someone pointed out to me that it might be IPV6 related.. if i find out more ill update this post.


[update 2]
: what happens in 0.4.12 is if one server interrupts it blocks the entire queue
that does seem to be the case, i put the "dead" server as backup for now (so i dont loose the info its just for the test) and it seems to be stable now and not crash..
not sure if it only applies to ipv6 or v4 as well (as the topic starter has).

[ZeroGhost]

i did hope its fixed in there