[Support] ich777 - Application Dockers


ich777

Recommended Posts

25 minutes ago, HpNoTiQ said:

Here you are :

When I'm connecting to VNC within my browser (whether chrome or Jdownloader), I only get Jdownloader webui. (or the 1st container i Launched)

Oh, now I get it, since you are using it through one container (one network) you also have to change the RFB port because they use both inside the container 5900 and two times 5900 can't work, I would suggest that you create a Variable with the Key: 'RFB_PORT' in both containers and as Value for the jDownlaoader2 '5905' and in the Chrome container '5906' (you don't have to create a port forwarding in your Wireguard container for that, this is only how they communicate internally in the container with noVNC and the VNC server).

 

Hope that makes sense.

Link to comment
7 minutes ago, EdB said:

I'd like to transition to LB from the user.script backup solution I'm currently using but without the email function it is not desirable for my requirement.

If I understand that correctly you want to get Mail notifications from LB?

I have to look into this but I think you have to configure it in LuckyBackup itself since you can't easily use the Mail notifications that are configured in Unraid.

Link to comment
7 minutes ago, ich777 said:

If I understand that correctly you want to get Mail notifications from LB?

I have to look into this but I think you have to configure it in LuckyBackup itself since you can't easily use the Mail notifications that are configured in Unraid.

Correct.  I'd like to use it, if available.  Is another method being used elsewhere or are people not sending status emails for jobs?

Link to comment
2 hours ago, ich777 said:

Oh, now I get it, since you are using it through one container (one network) you also have to change the RFB port because they use both inside the container 5900 and two times 5900 can't work, I would suggest that you create a Variable with the Key: 'RFB_PORT' in both containers and as Value for the jDownlaoader2 '5905' and in the Chrome container '5906' (you don't have to create a port forwarding in your Wireguard container for that, this is only how they communicate internally in the container with noVNC and the VNC server).

 

Hope that makes sense.

Perfectly working !
You should add it on github or docker as env variable!

Link to comment
58 minutes ago, HpNoTiQ said:

Perfectly working !
You should add it on github or docker as env variable!

Yes and no...

I dodn't added it because you actually can make things worse if you don't know what you are doing...

I have many other variables integrated in Containers that are not visible to users because these are "in case someone needs this for whatever reason" variables for me...

 

3 hours ago, EdB said:

Is another method being used elsewhere or are people not sending status emails for jobs?

No one ever requested this but I will look into it ASAP. ;)

  • Like 1
Link to comment
On 3/26/2021 at 5:28 PM, runraid said:

Yeah but what I'm concerned about is the fact that something was changed from underneath me. I didn't make these changes but suddenly there's this https://myunraid.ru/ showing up in my unraid templates. My template completely changed, seemed to be changed by a 3rd party actor.  These templates are weird - https://github.com/Muwahhidun/unraid-docker-templates and something nefarious seems to be going on.  The githistory on that repo is very recent. Yet I was using the VPN template for two years. So how did my unraid switch to a new template? 

 

Let me know if this thread isn't the right place to talk about this. I will switch to your template (Thanks for providing it!)

Are you perhaps running the Fix Common Problems plugin? Today I received the following warning from the plugin:

 

image.thumb.png.9846ba3e19715810548fe701e25ebb8c.png

 

For some reason Fix Common Problems thinks that the template URL is missing and is proposing the template of "Muwahhidun". I have no idea why it is doing this. It might be malicious, but I don't know enough about Unraid templates (yet) to give my opinion on this.

Link to comment
On 4/4/2021 at 10:27 AM, EdB said:

Correct.  I'd like to use it, if available.  Is another method being used elsewhere or are people not sending status emails for jobs?

I have now integrated 'sendemail' to the container and updated it in a few other different ways (Please be sure to grafik.png.f08fe37a506fc1b8e25edc21e9a26386.png the Container on the Docker page before trying this).

 

To send a mail configure luckyBackup like this:

Bildschirmfoto_2021-04-06_13-27-10.thumb.png.69c6527140d86d546fe7e5f351f7ba04.png

 

Please notice that I have appended '-xu' (SMTP username) and '-xp' (SMTP user password) to the arguments to actually being able to send an email because my mail server, as I think most mailservers nowerdays, need a authentication and it works without a problem (if you entered 'sendemail' at command click on the green arrow on the right side and it will fill in all the necessary arguments in the arguments line without the two arguments mentioned above).

 

Please report back if this is also OK for you because I don't have a easy way of send a Mail through a Docker container from Unraid.

 

  • Like 2
Link to comment

Thank you for the quick work!  Unfortunately, it didn't work and I think it's because it uses the default port 25.  is there a VAR for port selection?  smtp.live.com uses a different port (587) and it'd be more configurable for other servers or relays using a VAR for the port.  TLS was used during the attempt and this is definitely further than initially tested.  Thank you so much for this effort!

image.png.d10ae2f81e9fc6bf04dfcdc0d3ce5006.png

 

image.png.623d3f39ee20dc620a50b8de88cf3a5b.png

Link to comment

Same error.  After a little more investigation, now that I have confirmation that adding the port to the servername is functional, it might be a TCPIPv6 binding issue.  IF INET6 is indicating the correct error, that is.  Apparently, it's a common enough issue with docker where it will bind to v6 and if you're running v4 it will fail.  From what I've gone through so far (about 20 postings), there's no "easy" way to fix it and sometimes it's confined to an OS and docker combo (although this is muddled too because some say it works correctly on their same OS).  I'm not certain specifically, and will continue to scour.  My google-fu is still running validation but I thought I'd respond with the preliminary findings.

 

Thanks again!!!

Link to comment
6 hours ago, EdB said:

Same error.  After a little more investigation, now that I have confirmation that adding the port to the servername is functional, it might be a TCPIPv6 binding issue.  IF INET6 is indicating the correct error, that is.  Apparently, it's a common enough issue with docker where it will bind to v6 and if you're running v4 it will fail.  From what I've gone through so far (about 20 postings), there's no "easy" way to fix it and sometimes it's confined to an OS and docker combo (although this is muddled too because some say it works correctly on their same OS).  I'm not certain specifically, and will continue to scour.  My google-fu is still running validation but I thought I'd respond with the preliminary findings.

 

Thanks again!!!

Can you eventually try to use a different mail provider, just for testing purposes.

Eventually you have a Gmail account or something similar.

In my case 'sendemail' detected that it has to use port 587 and STARTLS and sent the mail successfully.

Link to comment
42 minutes ago, ich777 said:

Can you eventually try to use a different mail provider, just for testing purposes.

Eventually you have a Gmail account or something similar.

In my case 'sendemail' detected that it has to use port 587 and STARTLS and sent the mail successfully.

I had already tried it, it's the same INET6 error.  It also doesn't automatically select the correct port, it tries 25 by default, I suspect this is because it can't get an outbound connection.  I added the port to the server to try and it failed as well.  I tried gmail and my own personal cloud mail server without success.  From what I can tell, I have my OS network settings to use v4 Only but the docker is trying to use v6 as a default (probably the default lo for v6 that Unraid uses is causing it to appear as an available protocol).  My next test is to switch the network settings to use v4 and/or v6 and retest but I can't shutdown VM and docker right now.  Adding that SHOULD provide the ability for it to bridge to v4 (hybrid approach) but I'm uncertain how complex the Unraid failover configuration works.  I'd rather turn on v6 for the server than to remove routes.  I'm thinking that will work, or at least give a further error beyond this one.  I should be able to make the switch in about 10 hours and then I'll send an update.  Thanks!!!

Link to comment
33 minutes ago, EdB said:

It also doesn't automatically select the correct port, it tries 25 by default, I suspect this is because it can't get an outbound connection.

Eventually, in my case I haven't specified a port and it worked just fine where my ISP doesn't support connections over port 25.

 

33 minutes ago, EdB said:

My next test is to switch the network settings to use v4 and/or v6 and retest but I can't shutdown VM and docker right now.

That's really strange and interesting, because I also only use v4 and it is working without a problem.

 

33 minutes ago, EdB said:

I should be able to make the switch in about 10 hours and then I'll send an update.  Thanks!!!

No pressure on that, really appreciate the responses and the updates about your findings. :)

  • Like 1
Link to comment

Hello,

by default, data dir path is bind to /mnt/cache/appdata/photoprism. I changed that but I don't know why, it was reseted. So It has created a new "cache" folder in /mnt because my cache pool are not named cache. It would be great to use /mnt/user/appdata/photoprism because it will use the right cache pool assigned to the "appdata" share 

Link to comment
7 minutes ago, dada051 said:

by default, data dir path is bind to /mnt/cache/appdata/photoprism. I changed that but I don't know why, it was reseted.

Exactly. In my templates it is not possible that it changed autmoatically.

 

7 minutes ago, dada051 said:

It would be great to use /mnt/user/appdata/photoprism because it will use the right cache pool assigned to the "appdata" share

I do it that way because some applications inside Docker containers needs the "real" path to the directory where the files in otherwise they won't work and I also won't change that in my templates.

 

As said above it is very unlikely that this path is reseted out of nowhere because I have disabled to auto update user templates in my templates if I change something in the main template.

 

@Squid can you help here or imagine what changed the path?

Link to comment
7 minutes ago, dada051 said:

So I have to rename my cache pool to avoid any problems later ?

No, if you change the path inside the template so that it matches your configuration it won't change automatically.

 

EDIT: To what path did you set it?

Link to comment
15 minutes ago, dada051 said:

But It has been reset after that change (maybe after the docker update, but I'm not sure).

What change? Let's wait for Squid, eventually he knows why this happened.

As said above this shouldn't happen when you changed it once.

Link to comment

Maybe I'm wrong or I forgot something that I've done, but I changed the "data dir' path in the photoprism docker settings and this morning, I found that it is set to the default one. One thing I am almost sure is that using /mnt/user/ instead of /mnt/cache seems to work well. But it should need more testing 

Link to comment

Here's how the system handles path mappings.

 

On any given app, if there is a container path of /config, it doesn't matter what the template author puts into it as a host path by default.  Unraid will always override the host path with whatever the setting is for default appdata path (Settings - Docker) and append the name of the app after it.

 

On any given app, if there are any paths pre-populated (other than /config) in the template, CA will automatically adjust them IF they directly refer to an array drive or a cache pool.  It looks at the user's system, and if the template says "/mnt/cache/data" and there is no pool named "cache", it will change the path to be the next available cache pool (alphabetically).  If the user has no cache pool at all, then it will change the path to be /mnt/disk1/data (assuming there is a disk 1, otherwise it will change it to disk 2 etc)

 

These changes only happen when installing a new template and are the default settings.  If you're editing an existing app, then you can override the settings for /config all day long.  (Or override the paths when installing a new app - these are the default paths the system offers up to the user.)

 

As an aside, technically on 6.9+ there should be zero reason to ever still have to use /mnt/cache/appdata/... instead of /mnt/user/appdata/...  If the appdata share only exists on the cache-pool, the OS dereferences /mnt/user internally to instead be /mnt/cache/...

 

But, like everything else IF there are still problems with /mnt/user/appdata with certain applications, let Tom know.

Link to comment
7 hours ago, Squid said:

Here's how the system handles path mappings.

Thanks for the explanation. :)

 

But what about that:

20 hours ago, dada051 said:

by default, data dir path is bind to /mnt/cache/appdata/photoprism. I changed that but I don't know why, it was reseted.

 

 

7 hours ago, Squid said:

As an aside, technically on 6.9+ there should be zero reason to ever still have to use /mnt/cache/appdata/... instead of /mnt/user/appdata/...  If the appdata share only exists on the cache-pool, the OS dereferences /mnt/user internally to instead be /mnt/cache/...

Also tested it with one of the "problematic" game server Docker (PVK-II) and it won't startup successfully if the path is '/mnt/user/appdata/...' when the download finished it crashes and restarts again.

 

I will contact Tom @limetech about that when I finished testing with other "problematic" game server Dockers (TeamFortress 2, Don't Starve,...).

Link to comment

@ich777

 

Thanks very much for the owncast docker this is really interesting. A couple of questions

 

1. port 8080 is already being used for another docker. is it safe to just change this to another port in the template?

2. Is there a simple proxy-conf I can edit to use this docker with swag so I can use my own domain to reach the site?

Edited by enigma27
  • Like 1
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.