[Support] ich777 - Application Dockers

[ich777]

2 hours ago, Cornflake said:

hey @ich777 any reason Photoprism was moved to deprecated on your github repo ?

I'm not using it anymore and for me it's a pain to maintain because I have to research almost everything if someone has an issue with it.

What's really bugging me is that the container is outside of my control and I can't help much if something needs to be change don how the container is working (I'm not very comfortable using a container that someone else is maintaining).

 

However if someone wants to maintain it fell free to use the existing template that I have in my CA App repository on GitHub.

[mattw]

I just updated my LuckyBackup container from the previous release and now I am having authentication problems.  I saw your not and walked thru this process again thinking the update may need to have it done again...  https://forums.unraid.net/topic/83786-support-ich777-application-dockers/page/34/#comment-984853

 

This did not help...  when I start one of my existing jobs I get the following screen...

 

The problem appears to be that I am not able to auth to the remote system... What should I do to fix this?

[ich777]

1 hour ago, mattw said:

The problem appears to be that I am not able to auth to the remote system... What should I do to fix this?

Please read the third recommended post from this thread.

You have to put your keys in /luckybackup/.ssh not in /root/.ssh

[mattw]

I did that part, but did not copy the key back over and add it to authorized_keys.  I have it working now...  albeit with a newly generated key pair.

 

Thanks

[ich777]

1 hour ago, mattw said:

albeit with a newly generated key pair.

This is necessary because you most certainly had it in /root/.ssh correct? This will not be necessary next time if you leave it in /luckybackup/.ssh

 

The step of creating a new key pair is not necessary, the container does that on it's own on the first start or each start no key pair is found, the tutorial that I've linked is all you need to do.

[dataslayer]

I recently started seeing this error in LuckyBackup:

 

WebSocket server settings:
  - Listen on :8080
  - Web server. Web root: /usr/share/novnc
  - No SSL/TLS support (no cert file)
  - Backgrounding (daemon)
No protocol specified
qt.qpa.xcb: could not connect to display :0
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "" even though it was found.
This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, offscreen, vnc, xcb.

/opt/scripts/start-server.sh: line 86:   116 Aborted                 /usr/bin/luckybackup
 

 

The GUI doesn't come up but the software still seems to run. What can I do to fix this?

[ich777]

53 minutes ago, dataslayer said:

The GUI doesn't come up but the software still seems to run. What can I do to fix this?

On which Unraid version are you?

[dataslayer]

23 hours ago, ich777 said:

On which Unraid version are you?

I’m up to date - 6.12.4

[mattw]

So, I really love LuckyBackup!

 

I have an old Unraid server as my backup catcher, this machine sits turned off all of the time.  From the Unraid console I can issue the command etherwake -D bc:5f:f4:86:57:73 and wake the system up.  But, when I add the same command or etherwake -D -i br0 -b bc:5f:f4:86:57:73 from within the LuckyBackup Also Execute Before command it fails because it cannot find etherwake, same from the LuckyBackup console. 

 

I assume this is due to the fact that LB is running in a container that does not have etherwake installed and the container is not able to run commands on the OS directly.  How do I get around this?

 

[ich777]

39 minutes ago, dataslayer said:

I’m up to date - 6.12.4

This is really strange, have you yet tried to completely remove the container and then reinstall it from the CA App, it seems like your Docker layers for this container became corrupt.

[ich777]

3 minutes ago, mattw said:

I assume this is due to the fact that LB is running in a container that does not have etherwake installed and the container is not able to run commands on the OS directly.  How do I get around this?

You can write a script on the host that you can name however you want, lets name it "lb-etherwake.sh" with the following contents:

#!/bin/bash
apt-get update
apt-get -y install etherwake

 

...after that go to the template, click on "Add another Path, Port, Variable, Label or Device", as config type left it at Path, the container path must be: /opt/scripts/user.sh and on the host you have to go to the path where the script is located and you then have to manually append "lb-etherwake.sh" to the path, this will mount the script in the container.

It should look something like that:

Click on "Add" and then on "Apply", by doing all of that the script should pull etherwake on container start and it should be available in the container after you did all of that.

 

 

Please keep in mind that you maybe also have to change the network type from the container to Host since otherwise etherwake will most certainly not work (I have never tried that but if you set the network type to host in the template it should work because if you leave it on bridge your WOL package will be only be sent into the Docker subnet and never reach anything outside that network).

[mattw]

3 hours ago, ich777 said:

...after that go to the template, click on "Add another Path, Port, Variable, Label or Device", as config type left it at Path, the container path must be: /opt/scripts/user.sh and on the host you have to go to the path where the script is located and you then have to manually append "lb-etherwake.sh" to the path, this will mount the script in the container.

It should look something like that:

 

 

 

 

Trying to make sense of this section...  so is the script that I create located within the container or do I create it in the console of the main server in the cache directory? 

 

I manually did the update and install and it worked, but of course after a restart of the container it was gone.

 

 

[mattw]

So, in the process of trying to convert the container to host based, I have found what I think is a problem...

Here are my settings:

 

Have also tried the default of Privileged: off.

 

When I try to browse to the webgui, LB tries to take me to this address:

Notice the URL, wrong port.  Log indicates the wrong port as well and I cannot get it to change.

 

WebSocket server settings:
  - Listen on :8080
  - Web server. Web root: /usr/share/novnc
  - No SSL/TLS support (no cert file)
  - Backgrounding (daemon)
---Ensuring UID: 99 matches user---
---Ensuring GID: 100 matches user---
---Setting umask to 0000---
---Checking for optional scripts---
---No optional script found, continuing---
---Checking configuration for noVNC---
Nothing to do, noVNC resizing set to default
Nothing to do, noVNC qaulity set to default
Nothing to do, noVNC compression set to default
---Starting cron---
---Taking ownership of data...---
---Starting...---
---Preparing Server---
---ssh_host_rsa_key keys found!---
---ssh_host_ecdsa_key found!---
---ssh_host_ed25519_key found!---
---Starting ssh daemon---
---Resolution check---
---Checking for old logfiles---
---Starting TurboVNC server---
---Starting Fluxbox---
---Starting noVNC server---
---Starting ssh daemon---
---Starting luckyBackup---

 

Once this fails, I cannot switch back to bridge mode, the port setting seems to be stuck.  In a browser that was not aware of previous connections to scrutiny on port 8080 I do indeed get the LB GUI.  I have found that if I remove the container and reinstall it all is good again.

[alturismo]

50 minutes ago, mattw said:

Notice the URL, wrong port.  Log indicates the wrong port as well and I cannot get it to change.

 

when running in host mode, your port mappings are obsolete as you run in host mode

 

if you want to change the Port, may search this thread for howto ... been posted 1 - 2 pages back (again)

 

also, mapping in bridge 9900 <> 9900 wont work without changing the nativ app port ... consider this like a router port forwarding (when in bridge), but just changing the forwarding wont change the nativ listening port ... thats why you fail when you change back to bridge with this setting, port 9900 is just not listening ... there are 2 Variables you can add to change the nativ ports ... but then also you need to match your mappings ...

[te5s3rakt]

anyone having issues with the edge container. I just updated to the last update, and now web I browse to the Web UI, I just land on the VNC screen and the error "Failed to connect to server" top of the page.

Also opening the log window, the window pops up, then immediately closes

[ich777]

1 hour ago, te5s3rakt said:

anyone having issues with the edge container. I just updated to the last update, and now web I browse to the Web UI, I just land on the VNC screen and the error "Failed to connect to server" top of the page.

Also opening the log window, the window pops up, then immediately closes

Stop the container open the log and then post the contents of the log window.

[dataslayer]

22 hours ago, ich777 said:

This is really strange, have you yet tried to completely remove the container and then reinstall it from the CA App, it seems like your Docker layers for this container became corrupt.

I did try to uninstall and reinstall. But let me make sure I did it right:

I first go to the docker tab and click uninstall. I made sure that the check box was checked to fully remove. 

After the uninstall, I go back to the Apps tab and install the app again. 

 

If those are the right steps and there isn’t any clean up that needs to happen, then I did it correctly. Even after doing that, I’m getting that error. 

[ich777]

1 hour ago, dataslayer said:

I first go to the docker tab and click uninstall. I made sure that the check box was checked to fully remove. 

Yes.

 

1 hour ago, dataslayer said:

After the uninstall, I go back to the Apps tab and install the app again. 

I would also recommend that you remove the directory that was created for luckyBackup in the CA App.

[dataslayer]

6 hours ago, ich777 said:

I would also recommend that you remove the directory that was created for luckyBackup in the CA App.

That bit was the trick to flush out the configs. I had mistakenly thought that a remove of the app would remove the appdata. 

 

Now that I have completely removed it, it now starts up and I can get to the VNC version. 

 

I also had a backup from the day I posted this note, which all i cared about was the profiles, logs and snaps so that everything was as it was from the last "good" backup. 

 

I think I'm back in business. The key was the appdata folder removal too. Thanks ich777!

[Rusty6285]

On 10/27/2023 at 11:21 PM, dataslayer said:

That bit was the trick to flush out the configs. I had mistakenly thought that a remove of the app would remove the appdata. 

 

Now that I have completely removed it, it now starts up and I can get to the VNC version. 

 

I also had a backup from the day I posted this note, which all i cared about was the profiles, logs and snaps so that everything was as it was from the last "good" backup. 

 

I think I'm back in business. The key was the appdata folder removal too. Thanks ich777!

I'm having this same issue where luckybackup is unresponsive after last update. Same log errors as @dataslayerposted. I don't think I have a recent backup of my profiles... is there a way for me to roll back to the prior version so I can save my profiles, before removing/cleaning my config?

 

Edit - By copying out the contents of my 'profiles' folder inside the appdata, then removing both the image and the appdata folder for a reinstall, luckybackup is now working again on a fresh install...

 

Edit 2 - placing the saved profile files back in the new appdata seems to have restored my profiles, however they all are requiring a manual fix to the path, since it's doubled up on 'user/user' where it should only be once...

 

Edit 3 - I seem to have lost the ability to map to remote shares? they are correctly mounted but the path /remotes/ is completely absent in the interface as of this last update.

 

Edit 4 - I found the solution to my profile issues in edit 2 & 3 - seems in my prior install I had 'Container Path: /mnt/user' mapped to just '/mnt/' instead of '/mnt/user' to allow me access to remote shares. Mapping it back to '/mnt/' solved the double user issue, and allowed me to access remotes again.

 

Edited October 30, 2023 by Rusty6285

[ramonstohr]

Hi,

i updated your container for SABnzbd to the latest version today.

i get following error and no file gets unpacked:

Your UNRAR version is 0.00, we recommend version 5.50 or higher.<br />

 

What can i do about that?

 

Thanks and good evening

[ich777]

1 hour ago, Rusty6285 said:

I'm having this same issue where luckybackup is unresponsive after last update.

What do you mean with unresponsive? Have you cleared the browser cache?

 

1 hour ago, Rusty6285 said:

Edit 4

But I assume everything is now working correct?

[ich777]

1 hour ago, ramonstohr said:

What can i do about that?

Are you really sure that you are using my container and that you are on the latest version from the container itself? I don't see that over here, actually I see that:

[ramonstohr]

 

Yeah, dont know what happened?

 

[ich777]

26 minutes ago, ramonstohr said:

Yeah, dont know what happened?

Do you have custom scripts installed or something like that?

 

Please try to click force update and see if the issue is persists.