0 B Free Space [SMB mount and Linux]

[jakadia]

Hello everyone, I typically try and do some searching for my issue before I post in a forum, but this time I am having no luck finding any resolutions.

My current setup is running SABnzbd on a docker VM and on a separate VM I have Unraid running that manages all my storage. I have 34TB of free storage on the Unraid share that I share with my SABnzbd docker container and recently I have been getting told that I have 0 B of free space and all my downloads pause.

I can console into the SABnzbd container and I can see that the Unraid share is mounted and I can access all the files which tells me the share is mounted and working correctly. I also have Plex running on a separate container and that has not experienced any issues. I can change my completed and incomplete storage locations to a local folder and after saving the error disappears and shows the correct amount of free storage. I feel the problem lies around the mounted share and how it reports free space to SABnzbd.

Let me know if you need any logs or config setup info to help troubleshoot. All help will be appreciated.

Thanks!

[sander]

Indeed: HOW it reports.

I'll check how sabnzbd via a python call checks free system

BRB

[jakadia]

I appreciate you taking a look. The weird part is that it worked for over a year and now it just unexpectedly stopped. It would make more sense if I was tinkering or something but I literally haven’t done anything.

Maybe an update to Unraid or SABnzbd?

Thanks
Jake

[sander]

OK.

This is how SABnzbd does it: https://github.com/sabnzbd/sabnzbd/blob ... m.py#L1059

Can you do this within your docker, within the SABnzbd directory

sander@brixit:~$ cd git/sabnzbd/ # go to directory with SABnzbd.py

What does OS say, for working directory:

Code: Select all

$ df -h .
Filesystem      Size  Used Avail Use% Mounted on
/dev/nvme0n1p2  129G  119G  3,4G  98% /

What does SABnzbd say:

home dir:

Code: Select all

sander@zwart2204:~/git/sabnzbd$ python3 -c "import sabnzbd.filesystem as filesystem; print(filesystem.diskspace_base('/home/sander/'))   " 
(128.7086944580078, 3.374744415283203)

Another drive:

Code: Select all

$ python3 -c "import sabnzbd.filesystem as filesystem; print(filesystem.diskspace_base('/media/sander/spare2/'))   " 
(103.91062545776367, 93.91534805297852)

No dir input

Code: Select all

$ python3 -c "import sabnzbd.filesystem as filesystem; print(filesystem.diskspace_base(''))   " 
(0.0, 0.0)

Wrong input

Code: Select all

$ python3 -c "import sabnzbd.filesystem as filesystem; print(filesystem.diskspace_base('kjldskjldsfklj'))   " 
(0.0, 0.0)

[jakadia]

So after looking through my docker logs, I was able to find this error - validate protocol negotiate failed: -11

I ended up mounting my "media" folder on the docker host and mapping that to my SABnzbd container. Prior to this, I did the mounting inside the container. All seems to be working now.

[sander]

So after looking through my docker logs, I was able to find this error - validate protocol negotiate failed: -11

If I google "validate protocol negotiate failed: -11" it gives hits with SMB / CIFS. So filesystem (mount) problem. Indeed you should solve that; not a SABnzbd problem.

Good that it works again.

[jakadia]

I appreciate you taking time out of your day to help me. Have a great week!

[sander]

@safihre

I find defaulting to 0.0 confusing.

Have we considered returning float('nan') in case of problems? Or just -1? So that we can recognize.

Code: Select all

>>> x = float('nan')

>>> x == x
False

>>> x < 1
False
>>> x >= 1
False

[ckp33]

i have the same issue out of nowhere it shows i have 0 B free space even though i have plenty and other download tools to the same mounted smb share work fine. sabnzbd on ubuntu 20.04 mounted to a windows smb share seems to have started this problem only after upgrading to sab 4.2.3
hope someone can help. all my downloads will fail with not enough disk space error and pause. it was working for years until now.
i'll start a new topic on this since mine might be different.

[cactusCoffee4649]

Chiming in to say that I'm also having the same issue. Based on logs, the problem started on 4/14 after my server rebooted.

2024-04-14 02:35:39,231::INFO::[notifier:142] Sending notification: Error - Cannot read Watched Folder /complete-downloads (type=error, job_cat=None)
2024-04-14 02:35:39,231::ERROR::[dirscanner:180] Cannot read Watched Folder /complete-downloads
2024-04-14 02:35:39,233::INFO::[notifier:142] Sending notification: Warning - /incomplete-downloads is not writable at all. This blocks downloads.
To prevent all helpful warnings, disable Special setting 'helpful_warnings'. (type=warning, job_cat=None)
2024-04-14 02:35:39,233::WARNING::[misc:94] /incomplete-downloads is not writable at all. This blocks downloads.
To prevent all helpful warnings, disable Special setting 'helpful_warnings'.
2024-04-14 02:35:39,943::WARNING::[misc:94] /complete-downloads is not writable at all. This blocks downloads.

I'm using SAB in Docker on a PC running Ubuntu Server 22.04.4 LTS, and the download location is to a mounted Windows SMB share on a Synology NAS. I'm also getting the same issue in a jDownloader 2 container (no disk space available error when attempting to download), but oddly am not seeing the problem in my Firefox or qBittorrent containers.

[safihre]

SABnzbd can't reach it.. This isn't a SABnzbd problem.

[QuirkyKirkHax]

The 0B issue seems to be caused by a kernel update on Ubuntu 22.04:

(I would post a link, but new users are not allowed...It's bug #2060780 on the Ubuntu Launchpad site.)

However, there is an even more recent kernel update that appears to fix this now.

I am no longer seeing the 0B free issue now, but for some reason the SAB docker container is defaulting to local folders instead of the CIFS bind mounts originally assigned to the container...I even created a second dummy container to start fresh, and it seems unable to mount cifs, even though the OS now seems to be able to use it without issue.

Anyone else who experienced the 0B issue able to confirm?

[QuirkyKirkHax]

Digging a little deeper, I can see the container can see the cifs share without trouble, and they even appear to be mounted properly within the container....

Code: Select all

quirk@rainsvr:~$ docker exec -it sabnzbd bash
root@3156e87bdf70:/# ls
app  command  defaultsdocker-mods  downloadshome      init  lsiopy  mnt  package  rootsbin  sys  usr
bin  config   devdonate.txt   etcincomplete-downloads  lib   media   opt  proc  runsrv   tmp  var
root@3156e87bdf70:/# df -h
Filesystem                         Size  Used Avail Use% Mounted on
overlay                            244G  106G  128G  46% /
tmpfs                               64M     0   64M   0% /dev
shm                                 64M     0   64M   0% /dev/shm
/dev/mapper/ubuntu--vg-ubuntu--lv  244G  106G  128G  46% /config
//192.168.1.68/Movies And TV        35T   19T   17T  54% /downloads
tmpfs                              3.9G     0  3.9G   0% /proc/acpi
tmpfs                              3.9G     0  3.9G   0% /proc/scsi
tmpfs                              3.9G     0  3.9G   0% /sys/firmware
tmpfs                              3.9G     0  3.9G   0% /sys/devices/virtual/powercap

But when I try to do a pull with Sab, the files never show in the share folder, and fail to move properly. So I checked the local Download directory that Sab keeps creating for some reason, and all the files are in there...Its like it has two sets of download folders somehow.

This is the local thing I'm seeing...

Code: Select all

root@51b37612dac2:/# ls /config/Downloads
complete  incomplete

Threw some junk on the cifs share to check the ones in the root folder

Code: Select all

root@51b37612dac2:/# ls downloads
TESTCOMPLETE
root@51b37612dac2:/# ls incomplete-downloads/
TESTINCOMPLETE

But logging into the container with bash, the root download and incomplete download directories appear to be the share...I have no clue where the local ones are coming from and where they are being linked in the container.....

And deleting the ones in /config doesn't appear to help, since they regenerate on every reboot....

[QuirkyKirkHax]

Fuck Im an idiot....Forgot you could set the folder within webgui....

[cactusCoffee4649]

Digging a little deeper, I can see the container can see the cifs share without trouble, and they even appear to be mounted properly within the container....

Code: Select all

quirk@rainsvr:~$ docker exec -it sabnzbd bash
root@3156e87bdf70:/# ls
app  command  defaultsdocker-mods  downloadshome      init  lsiopy  mnt  package  rootsbin  sys  usr
bin  config   devdonate.txt   etcincomplete-downloads  lib   media   opt  proc  runsrv   tmp  var
root@3156e87bdf70:/# df -h
Filesystem                         Size  Used Avail Use% Mounted on
overlay                            244G  106G  128G  46% /
tmpfs                               64M     0   64M   0% /dev
shm                                 64M     0   64M   0% /dev/shm
/dev/mapper/ubuntu--vg-ubuntu--lv  244G  106G  128G  46% /config
//192.168.1.68/Movies And TV        35T   19T   17T  54% /downloads
tmpfs                              3.9G     0  3.9G   0% /proc/acpi
tmpfs                              3.9G     0  3.9G   0% /proc/scsi
tmpfs                              3.9G     0  3.9G   0% /sys/firmware
tmpfs                              3.9G     0  3.9G   0% /sys/devices/virtual/powercap

But when I try to do a pull with Sab, the files never show in the share folder, and fail to move properly. So I checked the local Download directory that Sab keeps creating for some reason, and all the files are in there...Its like it has two sets of download folders somehow.

This is the local thing I'm seeing...

Code: Select all

root@51b37612dac2:/# ls /config/Downloads
complete  incomplete

Threw some junk on the cifs share to check the ones in the root folder

Code: Select all

root@51b37612dac2:/# ls downloads
TESTCOMPLETE
root@51b37612dac2:/# ls incomplete-downloads/
TESTINCOMPLETE

But logging into the container with bash, the root download and incomplete download directories appear to be the share...I have no clue where the local ones are coming from and where they are being linked in the container.....

And deleting the ones in /config doesn't appear to help, since they regenerate on every reboot....

Hey, just wanted to say I'm noticing the exact same thing on my end. I successfully updated the kernel, but SAB is defaulting downloads to a Download folder in /config instead of the designated cifs share.

User Folders in the Web UI settings are showing the correct paths for complete and incomplete. Not sure what I have to do to get this working.