[[6.6.0] Docker CPU Pinning not persistent, params reappearing]

I can record a video of what's happening on my end if it would help since the diagnostic bundle apparently isn't showing the full story.  If there's anything specific (e.g. specific steps in the GUI, CLI commands) you'd like me to include, let me know.

 

I also tried manually creating the docker image via CLI and it's still adding the extra-param

root@undou:~# docker images
REPOSITORY                          TAG                 IMAGE ID            CREATED             SIZE
linuxserver/sonarr                  latest              7491f3da74a2        2 days ago          576MB
linuxserver/nzbget                  latest              7f1e8f590f23        2 days ago          75.7MB
linuxserver/tautulli                latest              df4f0f0d6ee9        9 days ago          188MB
brettm357/unifi                     latest              d259b09081bf        12 days ago         610MB
nanocurrency/nano-beta              latest              5d46b0c2ac89        13 days ago         122MB
nanocurrency/nano                   latest              cdecaab45885        13 days ago         122MB
adamfisher90/docker-deluge-1.3.11   latest              249049a9cda6        17 months ago       142MB
root@undou:~# /usr/local/emhttp/plugins/dynamix.docker.manager/scripts/docker create --name='radarr' --net='br0' --ip='10.0.0.77' -e TZ="America/New_York" -e HOST_OS="Unraid" -e 'TCP_PORT_7878'='7878' -e 'PUID'='99' -e 'PGID'='100' -v '/mnt/movies/.snatched':'/downloads':'rw' -v '/mnt/movies':'/movies':'rw' -v '/mnt/media/Anime':'/anime':'rw' -v '/mnt/user/appdata/radarr':'/config':'rw' 'linuxserver/radarr'

Even after saving the template update to remove the extra-param, I checked the actual template file and it persisted, meaning that template authoring mode isn't working correctly for some reason.  Due to this, I deleted the image/container once again and created it from an Unraid provided template that didn't have an extra-param defined (https://forums.unraid.net/topic/53758-support-linuxserverio-radarr/).

 

After creating this new instance, the CPU pinning extra-param resurfaced once again:

 

I had to manually go into "/boot/config/plugins/dockerMan/templates-user/my-Radarr.xml" and change "<ExtraParams>--cpuset-cpus=1</ExtraParams>" to "<ExtraParams/>", along with any other values I wanted to persist. 

 

Port mappings via config settings also aren't working correctly as it's using the legacy port mapping style (https://docs.docker.com/network/links/#environment-variables).  Even though I have the following in my ".../my-NZBget.xml" template, it's still translating to "-e 'TCP_PORT_6789'='6788'" upon creating the docker container with the effective port mapping of "-p 6789:6789" (which is incorrect).  I have to manually add in an extra-param of "-p 6788:6789", and remove the config entry for the port mapping for it to work correctly.

<!-- Snipped Code -->  
<Networking>
    <Mode>br0</Mode>
    <Publish>
      <Port>
        <HostPort>6788</HostPort>
        <ContainerPort>6789</ContainerPort>
        <Protocol>tcp</Protocol>
      </Port>
    </Publish>
</Networking>
<!-- Snipped Code -->
<Config Name="Host Port" Target="6789" Default="6788" Mode="tcp" Description="Container Port: 6789" Type="Port" Display="always" Required="false" Mask="false">6788</Config>

CPU pinning is also not saving in the template via the GUI.  I had to manually change "<CPUset/>" to "<CPUset>0,1</CPUset>" in the XML template.  After doing so, the GUI correctly shows the following when loading the template:

 

I ended up just creating my own repo of templates, which appears to work correctly.  The only thing that's still not working is port forwarding (i.e. '-p 6788:6789') even when it's in the extra params or even specifying '/tcp'.

 

[Bridge 0 Seems to be gone with 6.6]

4 hours ago, The_Doctor said:

Ran into this same issue. I had 2 plex dockers running, one for test, and one for Prod. My Prod one used br0 and a /24 IP and was set for use from outside. I thought something broke on the docker, but it turns out if i switch to use Host it worked just fine, but I cant do that with another plex docker using the same settings.

Just curious, but do you have 'preserve user-defined networks' enabled?  I don't and still had br0 enabled with the same settings after the upgrade.  You have to disable docker and turn on advanced mode to see these settings, BTW.

 

[[6.6.0] Docker CPU Pinning not persistent, params reappearing]

On 9/21/2018 at 4:28 AM, bonienl said:

Tested, and adding/removing "extra parameters" works as expected.

It is unclear why in your case it isn't working. Workaround: delete the container and recreate it

As I mentioned, I already tried deleting and recreating it and even tried using a new template.  It continues to add the parameters.  I think it's something screwy with the CPU Pinning options in Settings.  I just checked and there are values marked again in there that I never set.  In fact, I had removed those settings, saved, and rebooted to confirm.  It seems to have a mind of its own.

 

On 9/21/2018 at 10:19 PM, Squid said:

Are you positive that the template you're adding the app from does not already contain the cpu pinning in extra parameters?

Yes, I mentioned that I even created a new template that doesn't have extra parameters set, nor the GUI options.  As soon as I create it from the template it's adding the extra params.

 

[6.6 Upgrade - GPU Passthrough Issues]

You scared me a little as I came here trying to find info on the new PCIe ACS Override options (e.g. Downstream, Multifunction) without any luck, but just checked out my VM and GPU pass-through is still working after the upgrade.  One difference I noticed between our setups was the ACS Override settings...

 

Yours

Quote

BOOT_IMAGE=/bzimage pcie_acs_override=downstream,multifunction vfio-pci.ids=10de:1b80 vfio-pci.ids=10de:10f0 vfio-pci.ids=1b73:1100 vfio_iommu_type1.allow_unsafe_interrupts=1 initrd=/bzroot

 

Mine

Quote

BOOT_IMAGE=/bzimage pcie_acs_override=downstream initrd=/bzroot

 

I'm not sure if you manually added the other portions or not, but mine is much more simplistic.  Both your 6.5.3 and 6.6.0 installs appear to have the same settings, so maybe that's not the issue?  I'll attach my Diag bundle in case it helps you out.

 

undou-diagnostics-20180921-0213.zip

[Bridge 0 Seems to be gone with 6.6]

My br0 network persisted through the upgrade, but I can't recall if I manually created it or not in the first place.