Up until two days ago I was running SABnzbd (along with Couchpotatoe and Sickbeard) via Docker. 

 

All was working well until I started getting "unpacking failed, write error or disk is full" errors. I also seemed to have two different containers installed for SABnzbd.

 

I restarted my server to see if this would clear the issue, however on re-boot Docker wouldn't reload, meaning I had to delete the image and start again.  Luckily my config files remaind, so this only took 10 minutes or so.

 

The issue now is when SABnzbd downloads the files, it attempts to move them to the array but instead of moving them deletes them.  Also, SABnzbd fails to recognise my user share "Plex" and instead just reads from my first disk.

 

SABnzbd uses the following folder structure

Config > mnt/disk1/appdata/SABnzbd

Downloads > mnt/user/plex/ (made up of disk1 and disk2)

 

Has anyone come across this before?  Are these issues because I don't have a cache drive?

 

I've attached a copy of my SABnzbd log and also my servers diagnostics.

 

Please can someone advise what I need to do to fix this?

 

Cheers

mediaserver-diagnostics-20150913-1351.zip

Your SABnzbd logs weren't attached, but you might want to look at this series of posts and see if you've got Sab set up correctly.

 

Also, you should post a screen shot of the setting's for docker (in particular the section in the middle that shows the free space)

Sorry, the log wouldn't attach due to it being too big.  Nor could I add it as "code" in to my post.

 

I've half fixed it by changing the download directory to mnt/disk1/appdata/sabnzbd/downloads, however this will only allow me to have the downloads on one disk.  If I try to move them to the Plex User Share, my PC trys to download them to my machine before uploading them back to unRaid which will take hours.

 

By "settings for docker" do you mean this attached?

 

The docker IMG isn't full, (and I realized that I forgot to post the URL for the thread I wanted you to check out.)  http://lime-technology.com/forum/index.php?topic=42468.msg403951#msg403951

 

Post a snippet of the log from SAB showing that error, (bunch of lines before and after)

It's not full, but is at 11GB with only 4 apps installed in Docker.  Should 4 apps (plex, Sab, Couchpotatoe and Sickbeard) take up 11GB?

 

I'm not getting the disk full error anymore as I've reinstalled SAB.  Instead I'm just losing data, which I can't find in the SAB log.

 

Is it possible to set the config file for sab as mnt/user/appdata/SaBnzbd so that the config is saved across the array and not just to disk1?

I've managed to solve the issue.  I'm unsure if it's the correct way to do it or a bodge job?

 

I've pointed my SABnzbd config and downloads file to both use the user share /mnt/user/appsdata/sabnzbd and /mnt/user/appsdata/sabnzbd/downloads.

 

This now shows that I have 3TB free space remaining and no more missing downloads.

 

What will be the downsides of this?  I assume power consumption will be up due to all of the disks on the array spinning?  How can I get around this to do it another way?

 

Thanks

I've managed to solve the issue.  I'm unsure if it's the correct way to do it or a bodge job?

 

I've pointed my SABnzbd config and downloads file to both use the user share /mnt/user/appsdata/sabnzbd and /mnt/user/appsdata/sabnzbd/downloads.

 

This now shows that I have 3TB free space remaining and no more missing downloads.

 

What will be the downsides of this?  I assume power consumption will be up due to all of the disks on the array spinning?  How can I get around this to do it another way?

 

Thanks

If the appsdata share is set to be cache-only (which is what I would expect) then the paths above are directly equivalent to /mnt/cache/appsdata/sabnzbd and /mnt/cache/appsdata/sabnzbd/downloads.  All the downloads will be written to the cache drive, and no data drives will be spun up.

 

Hi,

 

I don't have a cache disk though? 

 

Saying that I've just logged in to unRaid Webui and only disk1 and parity are spun up, so it seems to be working OK.