[Support] Joshndroid's Docker Repo Support Thread

[own_rhino]

On 9/3/2023 at 1:44 PM, Ingenioes said:

I think the version on dockerhub is outdated, the official instance, that has been working for me (nitter.net) is on version 2023.09.02 while the dockerhub img is still on 2023.08.08.

 

On 9/3/2023 at 3:53 PM, Joshndroid said:

If you check the file ages, the default 'master' branch is a month behind, which is in line with the dockerhub image. They are using the 'guest_accounts' branch. My guess is that you would likely need to find a repo that is hosting a built image or build and host one yourself that you can import back into unraid. 

 

Ah great, that makes sense. I might build my own image based on the branch as a temporary solution.

[heckpiet]

Hi Joshndroid,

 

where did you put the .env for microbin?

I want to modify some of the settings.

But I cant find the configuration.

 

https://microbin.eu/docs/installation-and-configuration/configuration

 

Thanks

Piet

[Joshndroid]

31 minutes ago, heckpiet said:

Hi Joshndroid,

 

where did you put the .env for microbin?

I want to modify some of the settings.

But I cant find the configuration.

 

https://microbin.eu/docs/installation-and-configuration/configuration

 

Thanks

Piet

 

Within the docker page you should be able to add them as a 'variable' as it lists these as environment variables.

[heckpiet]

thanke mate

 

[cambriancatalyst]

Hi, this was working for the longest time and, at some point, it broke in my environment. Have tried re-pulling the image, redoing the config, host vs. bridge network modes, and rebuilding the DB but for some reason the container just finds itself in a constant state of restart. Here is what is getting run after I apply the docker template. Does this look right? Is there anything else I can attempt to troubleshoot? Any help or suggestions you can provide are very much appreciated. Thank you

 

```

docker run
  -d
  --name='invidious'
  --net='br0'
  --ip='192.168.1.27'
  -e TZ="America/New_York"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Tower"
  -e HOST_CONTAINERNAME="invidious"
  -e 'TCP_PORT_3000'='3000'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:3000]/'
  -l net.unraid.docker.icon='https://raw.githubusercontent.com/Joshndroid/joshndroid-unraid-docker-templates/main/icons/invidious.png'
  -v '/mnt/user/appdata/invidious/config/':'/invidious/config':'rw'
  --restart=always 'quay.io/invidious/invidious' 

69cd5458af0048cac4b74a1fda58a1652db5971dbb3199bbe07518db4de227b1

```

[flosken]

13 hours ago, cambriancatalyst said:

Hi, this was working for the longest time and, at some point, it broke in my environment. Have tried re-pulling the image, redoing the config, host vs. bridge network modes, and rebuilding the DB but for some reason the container just finds itself in a constant state of restart. Here is what is getting run after I apply the docker template. Does this look right? Is there anything else I can attempt to troubleshoot? Any help or suggestions you can provide are very much appreciated. Thank you

 

```

docker run
  -d
  --name='invidious'
  --net='br0'
  --ip='192.168.1.27'
  -e TZ="America/New_York"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Tower"
  -e HOST_CONTAINERNAME="invidious"
  -e 'TCP_PORT_3000'='3000'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:3000]/'
  -l net.unraid.docker.icon='https://raw.githubusercontent.com/Joshndroid/joshndroid-unraid-docker-templates/main/icons/invidious.png'
  -v '/mnt/user/appdata/invidious/config/':'/invidious/config':'rw'
  --restart=always 'quay.io/invidious/invidious' 

69cd5458af0048cac4b74a1fda58a1652db5971dbb3199bbe07518db4de227b1

```

Have you set the HMAC_key variable?

https://github.com/iv-org/invidious/issues/3854

[Joshndroid]

On 10/31/2023 at 12:36 PM, cambriancatalyst said:

Hi, this was working for the longest time and, at some point, it broke in my environment. Have tried re-pulling the image, redoing the config, host vs. bridge network modes, and rebuilding the DB but for some reason the container just finds itself in a constant state of restart. Here is what is getting run after I apply the docker template. Does this look right? Is there anything else I can attempt to troubleshoot? Any help or suggestions you can provide are very much appreciated. Thank you

 

```

docker run
  -d
  --name='invidious'
  --net='br0'
  --ip='192.168.1.27'
  -e TZ="America/New_York"
  -e HOST_OS="Unraid"
  -e HOST_HOSTNAME="Tower"
  -e HOST_CONTAINERNAME="invidious"
  -e 'TCP_PORT_3000'='3000'
  -l net.unraid.docker.managed=dockerman
  -l net.unraid.docker.webui='http://[IP]:[PORT:3000]/'
  -l net.unraid.docker.icon='https://raw.githubusercontent.com/Joshndroid/joshndroid-unraid-docker-templates/main/icons/invidious.png'
  -v '/mnt/user/appdata/invidious/config/':'/invidious/config':'rw'
  --restart=always 'quay.io/invidious/invidious' 

69cd5458af0048cac4b74a1fda58a1652db5971dbb3199bbe07518db4de227b1

```

 

Have you had a look in the logs? You sometimes can catch what is going on while watching it (especially if you are getting reboots where the logs wipe)

 

On 11/1/2023 at 2:02 AM, flosken said:

Have you set the HMAC_key variable?

https://github.com/iv-org/invidious/issues/3854

 

If OP has not added this, it would almost definitely be all/part of their issue.

It was not originally required until somewhat recently.

[Team_Dango]

Just installed the Baikal docker and I am hoping for advice on getting it correctly routed through NPM. The official baikal installation docs have an example Nginx config and mention needing to keep the "Specific" directory from being accessible. Is this handled by the docker container (it appears to have nginx internally), or do I need to add additional configuration in NPM? Thank you!

[Joshndroid]

On 12/5/2023 at 12:13 PM, Team_Dango said:

Just installed the Baikal docker and I am hoping for advice on getting it correctly routed through NPM. The official baikal installation docs have an example Nginx config and mention needing to keep the "Specific" directory from being accessible. Is this handled by the docker container (it appears to have nginx internally), or do I need to add additional configuration in NPM? Thank you!

 

I run mine with my own external NPM.

I use the host IP address and the port as my input into NPM.

 

[martial]

For people using microbin, I recommend modifying the "AppData" directory within the container to "/app" so that the files are not stored only within the container (and would not survive a container update, therefore)

 

Also, you can add environment variables as described in https://microbin.eu/docs/installation-and-configuration/configuration/

to enable things like
 

<Config Name="MICROBIN_ENCRYPTION_CLIENT_SIDE" Target="MICROBIN_ENCRYPTION_CLIENT_SIDE" Default="false" Mode="" Description="Enables server-side encryption.&#13;&#10;https://microbin.eu/docs/installation-and-configuration/configuration/" Type="Variable" Display="always" Required="false" Mask="false">true</Config>

<Config Name="MICROBIN_DISABLE_TELEMETRY" Target="MICROBIN_DISABLE_TELEMETRY" Default="false" Mode="" Description="Disables telemetry if set to true" Type="Variable" Display="always" Required="false" Mask="false">true</Config>

<Config Name="MICROBIN_ENCRYPTION_SERVER_SIDE" Target="MICROBIN_ENCRYPTION_SERVER_SIDE" Default="false" Mode="" Description="Enables client-side encryptio" Type="Variable" Display="always" Required="false" Mask="false">true</Config>

<Config Name="MICROBIN_PRIVATE" Target="MICROBIN_PRIVATE" Default="true" Mode="" Description="Enables private pastas" Type="Variable" Display="always" Required="false" Mask="false">true</Config>

<Config Name="MICROBIN_GC_DAYS" Target="MICROBIN_GC_DAYS" Default="90" Mode="" Description="Sets the garbage collector time limit. Pastas not accessed for N days are removed even if they are set to never expire. Default value: 90. To turn off GC: 0." Type="Variable" Display="always" Required="false" Mask="false">0</Config>

<Config Name="MICROBIN_ENABLE_BURN_AFTER" Target="MICROBIN_ENABLE_BURN_AFTER" Default="0" Mode="" Description="Sets the default burn after setting on the main screen. Default value: 0. Available expiration options: 1, 10, 100, 1000, 10000, 0 (= no limit)" Type="Variable" Display="always" Required="false" Mask="false">0</Config>

<Config Name="MICROBIN_PUBLIC_PATH" Target="MICROBIN_PUBLIC_PATH" Default="" Mode="" Description="Add the given public path prefix to all urls. This allows you to host MicroBin behind a reverse proxy on a subpath.&#13;&#10;You need to set the public path for QR code sharing to work." Type="Variable" Display="always" Required="false" Mask="false">https://yoururl/</Config>

<Config Name="MICROBIN_QR" Target="MICROBIN_QR" Default="false" Mode="" Description="Enables generating QR codes for pastas.&#13;&#10;This feature requires the public path to also be set." Type="Variable" Display="always" Required="false" Mask="false">true</Config>

<Config Name="MICROBIN_DEFAULT_EXPIRY" Target="MICROBIN_DEFAULT_EXPIRY" Default="24hour" Mode="" Description="Sets the default expiry time setting on the main screen. Default value: 24hour Available expiration options: 1min, 10min, 1hour, 24hour, 1week, never" Type="Variable" Display="always" Required="false" Mask="false">never</Config>

 

I am adding a copy of my modified XML file for those interested.

 

Hoping this helps some -- M

microbin.xml

Edited January 4 by martial

[Joshndroid]

On 1/4/2024 at 2:22 PM, martial said:

For people using microbin, I recommend modifying the "AppData" directory within the container to "/app" so that the files are not stored only within the container (and would not survive a container update, therefore)

 

~snip~

microbin.xml 3.92 kB · 2 downloads

thanks for the post

I'm happy to take a pull request to get the changes across if your interested in knocking one up

Edited January 13 by Joshndroid

[martial]

23 hours ago, Joshndroid said:

thanks for the post

I'm happy to take a pull request to get the changes across if your interested in knocking one up

Happy to do so. Please bear with me, as it might take a couple of days as I am out of town

Edited January 14 by martial

[Joshndroid]

On 1/14/2024 at 1:36 PM, martial said:

Happy to do so. Please bear with me, as it might take a couple of days as I am out of town

 

No dramas,

I have seen that it has come through and merged it, Thanks  

I have had the docker running for so long and never actually have needed to use it apart from a couple test runs so good to see someone uses it and gets some more functionality out of it

[martial]

On 1/16/2024 at 11:11 PM, Joshndroid said:

 

No dramas,

I have seen that it has come through and merged it, Thanks  

I have had the docker running for so long and never actually have needed to use it apart from a couple test runs so good to see someone uses it and gets some more functionality out of it

Yes, I was looking for a simple pastebin solution, without a DB, and that could do text and binaries ... and the simple word combination was useful. I tried it and started to wonder why the UI was missing some of the features from the webpage, so I checked

[olco]

I have been enthusiastically using Baikal as a Docker container for our family appointments for about a year. When I switched to a new smartphone, I realized that I could no longer log in with existing users.

I searched high and low but couldn't find a solution.

 

I can still log in as an admin, but not with the users via http://ip-adress/dav.php. Neither via DAVx5 nor via web browser does this work. My wife can still create and synchronize appointments, but she hasn't changed her smartphone in the meantime.

 

However, when I create a new user, logging in via browser and DAVx5 is possible without any problems.

[Mrtj18]

I would like to thank you so much for the invidious container, I had to read, read and re-read the comments and install instructions before attempting to install the container. But I got it working in less than an hour.

 

I would like to mention, if your new to this like me, the big hangup I was stuck on. It's the postgres install. When you initially download the container it is pre-filled with info for the database name, user, and database password. I had to remove all that and just use postgres for all the above on install. If not I couldn't login.

[deadmeu]

Hi, considering that Libreddit doesn't work and is no longer supported, would you consider replacing it with https://github.com/redlib-org/redlib?

Edited February 20 by deadmeu

[Joshndroid]

On 2/20/2024 at 7:01 PM, deadmeu said:

Hi, considering that Libreddit doesn't work and is no longer supported, would you consider replacing it with https://github.com/redlib-org/redlib?

 

On 2/13/2024 at 8:22 AM, olco said:

I have been enthusiastically using Baikal as a Docker container for our family appointments for about a year. When I switched to a new smartphone, I realized that I could no longer log in with existing users.

I searched high and low but couldn't find a solution.

 

I can still log in as an admin, but not with the users via http://ip-adress/dav.php. Neither via DAVx5 nor via web browser does this work. My wife can still create and synchronize appointments, but she hasn't changed her smartphone in the meantime.

 

However, when I create a new user, logging in via browser and DAVx5 is possible without any problems.

 

Honestly, I am not quite sure.

Mine has been working just fine for years now and only very recently (ie this week) I had an issue where it did not want to connect to the MariaDB database i was using with it. I tried troubleshooting myself with various things and the only thing that worked for me was a full nuke and i moved back to a sqlite3 database. Weird that after all this time it now decides to die. The only thing that we have in common i understand is the length of time of it working till it didn't but it taught me realistically that calendar backups external to the database are probably more important than backing up the database or db file itself.

 

What may work for yourself is exporting all of your calendars to an .ics file.. if everything you try doesnt work you can nuke it and just add your users again fresh then import your .ics files to the relevant calendar and keep going.

 

On 2/15/2024 at 2:47 AM, Mrtj18 said:

I would like to thank you so much for the invidious container, I had to read, read and re-read the comments and install instructions before attempting to install the container. But I got it working in less than an hour.

 

I would like to mention, if your new to this like me, the big hangup I was stuck on. It's the postgres install. When you initially download the container it is pre-filled with info for the database name, user, and database password. I had to remove all that and just use postgres for all the above on install. If not I couldn't login.

 

Invidious was the hardest one to get working by far... the most mucking about... but once i got it set up it has literally been the most stable and set and forget container I have ever used. Good to see others like it as well.

 

On 2/20/2024 at 7:01 PM, deadmeu said:

Hi, considering that Libreddit doesn't work and is no longer supported, would you consider replacing it with https://github.com/redlib-org/redlib?

 

I did see that earlier. I will have to see if its a fairly simple one to just copy straight over. I use troddit myself as i typically will login as well as a like their layout.

[Joshndroid]

On 2/20/2024 at 7:01 PM, deadmeu said:

Hi, considering that Libreddit doesn't work and is no longer supported, would you consider replacing it with https://github.com/redlib-org/redlib?

 

I have swapped out Libreddit for Redlib here. I was able to get it going fairly easily as essentially a drop in replacement at this time.

Let me know if there are any issues you see with the deploy.

https://github.com/Joshndroid/joshndroid-unraid-docker-templates/commit/17494dafc8a0d33e7a4203f104a82f85b7249737

[Mrtj18]

By the way if anyone else was stuck on the reverse proxy with ngix proxy mananger with the Invidious containter this is what i had to do....

 

-from the directions i had to set my external_port to 1443. ( I changed this from 443 on my setup, unraid was already on 443)

 

-the domain is set to my top level domain so its domain.com, not your subdomain.domain.com

 

-then the http_only line was set to true.

 

-then i restarted the ngix, the invidious container, then I had to clear my cookies in my browser. Because the invidious app would error out and not play any vidoes if you dont do this!

 

Then Bam! Profit

 

By the way the way the scheme is set to http in my ngix setup.

 

Hope this helps someone!

[Joshndroid]

55 minutes ago, Mrtj18 said:

By the way if anyone else was stuck on the reverse proxy with ngix proxy mananger with the Invidious containter this is what i had to do....

 

-from the directions i had to set my external_port to 1443. ( I changed this from 443 on my setup, unraid was already on 443)

 

-the domain is set to my top level domain so its domain.com, not your subdomain.domain.com

 

-then the http_only line was set to true.

 

-then i restarted the ngix, the invidious container, then I had to clear my cookies in my browser. Because the invidious app would error out and not play any vidoes if you dont do this!

 

Then Bam! Profit

 

By the way the way the scheme is set to http in my ngix setup.

 

Hope this helps someone!

 

I am not sure how you have various other things setup in your unraid system however I also use Nginx Proxy Manager and i don't have to do/use any of these steps but good to know there is another method for those if they experience a similar issue

[Joshndroid]

I have removed Nitter from both this thread as well as from the repo.

 

The project has been on its knees for a while and has been declared dead by the author here a little while back - https://github.com/zedeus/nitter/issues/1155#issuecomment-1913361757

 

That was one of the first templates that got me interested in making some templates for unraid, it had been a good ride.

 

At this time there appears to be no alternative that works and everything essentially goes through X now, similar to how there are no longer any insta or facebook front ends. RIP bibliogram.

 

There are active campaigns by google/youtube to restrict and break invidious instances. Thankfully at this time, invidious works fine for myself as well as the maintainers there being very active in keeping that project alive.

[serversonfire]

Once again I needed to install Invidious (should have never deleted it) and having trouble with NPM again.  Working fine locally.

 

I'm sure the database and all that is right so it comes down to the config file and NPM settings.

This is my config setting where I imagine I have to be doing something wrong.

 

In docker I changed port 3000 to port 3666 for the invidious setup, thats the only template change made.  This is on my proxynet network where all my other containers that I expose are, and they are all working with NPM no problem, so I'm positive I've misconfigured something specific to invidious.
 

# -----------------------------
#  Network (inbound)
# -----------------------------

##
## Port to listen on for incoming connections.
##
## Note: Ports lower than 1024 requires either root privileges
## (not recommended) or the "CAP_NET_BIND_SERVICE" capability
## (See https://stackoverflow.com/a/414258 and `man capabilities`)
##
## Accepted values: 1-65535
## Default: 3000
##
#port: 3666

##
## When the invidious instance is behind a proxy, and the proxy
## listens on a different port than the instance does, this lets
## invidious know about it. This is used to craft absolute URLs
## to the instance (e.g in the API).
##
## Note: This setting is MANDATORY if invidious is behind a
## reverse proxy.
##
## Accepted values: 1-65535
## Default: <none>
##
#external_port: 3666

##
## Interface address to listen on for incoming connections.
##
## Accepted values: a valid IPv4 or IPv6 address.
## default: 0.0.0.0  (listen on all interfaces)
##
#host_binding: 0.0.0.0

##
## Domain name under which this instance is hosted. This is
## used to craft absolute URLs to the instance (e.g in the API).
## The domain MUST be defined if your instance is accessed from
## a domain name (like 'example.com').
##
## Accepted values: a fully qualified domain name (FQDN)
## Default: <none>
##
domain: sub.ihavemydomain.here

##
## Tell Invidious that it is behind a proxy that provides only
## HTTPS, so all links must use the https:// scheme. This
## setting MUST be set to true if invidious is behind a
## reverse proxy serving HTTPs.
##
## Accepted values: true, false
## Default: false
##
https_only: true

##
## Enable/Disable 'Strict-Transport-Security'. Make sure that
## the domain specified under 'domain' is served securely.
##
## Accepted values: true, false
## Default: true
##
#hsts: true

 

 

 

Please and thank you!

[Joshndroid]

  

6 hours ago, serversonfire said:

Once again I needed to install Invidious (should have never deleted it) and having trouble with NPM again.  Working fine locally.

 

I'm sure the database and all that is right so it comes down to the config file and NPM settings.

This is my config setting where I imagine I have to be doing something wrong.

 

In docker I changed port 3000 to port 3666 for the invidious setup, thats the only template change made.  This is on my proxynet network where all my other containers that I expose are, and they are all working with NPM no problem, so I'm positive I've misconfigured something specific to invidious.
 

# -----------------------------
#  Network (inbound)
# -----------------------------

##
## Port to listen on for incoming connections.
##
## Note: Ports lower than 1024 requires either root privileges
## (not recommended) or the "CAP_NET_BIND_SERVICE" capability
## (See https://stackoverflow.com/a/414258 and `man capabilities`)
##
## Accepted values: 1-65535
## Default: 3000
##
#port: 3666

##
## When the invidious instance is behind a proxy, and the proxy
## listens on a different port than the instance does, this lets
## invidious know about it. This is used to craft absolute URLs
## to the instance (e.g in the API).
##
## Note: This setting is MANDATORY if invidious is behind a
## reverse proxy.
##
## Accepted values: 1-65535
## Default: <none>
##
#external_port: 3666

##
## Interface address to listen on for incoming connections.
##
## Accepted values: a valid IPv4 or IPv6 address.
## default: 0.0.0.0  (listen on all interfaces)
##
#host_binding: 0.0.0.0

##
## Domain name under which this instance is hosted. This is
## used to craft absolute URLs to the instance (e.g in the API).
## The domain MUST be defined if your instance is accessed from
## a domain name (like 'example.com').
##
## Accepted values: a fully qualified domain name (FQDN)
## Default: <none>
##
domain: sub.ihavemydomain.here

##
## Tell Invidious that it is behind a proxy that provides only
## HTTPS, so all links must use the https:// scheme. This
## setting MUST be set to true if invidious is behind a
## reverse proxy serving HTTPs.
##
## Accepted values: true, false
## Default: false
##
https_only: true

##
## Enable/Disable 'Strict-Transport-Security'. Make sure that
## the domain specified under 'domain' is served securely.
##
## Accepted values: true, false
## Default: true
##
#hsts: true

 

 

 

Please and thank you!

It would appear you need to remove the '#' from the port lines for external port and for port... both of those appear to have the '#' still there... 

[i_max]

Hello,

 

I've installed Invidious. I'm able to host it via NGINX Proxy Manager. However, I'm not able to set quality to dash (hd720 works), and I'm getting the "The media could not be loaded, either because the server or network failed or because the format is not supported." error.

 

I'm wondering if anyone else has this issue. I had followed the instructions to use postgres 12.5 as well as setting up the database as per original post. Since everything else works, I'm assuming most things are set correctly. Here is snippets of my config file which I'm not 100% sure about:

 

https_only: true

external_port: 18443

hsts: true
disable_proxy: false

force_resolve:
port: 3000

host_binding: 0.0.0.0
pool_size: 100

 

The proxy video option, doesn't tend to do anything in the settings in the GUI wrt the error above. I'm really trying to get the videos to play in 1080. To add a few more details, I've Pihole setup in the network. In the network port 80 is blocked, and invidious has Authelia authenticating before being directed to the gui.

 

 

 

Edit: It was fixed I had to set the port to 443. (required a few more restarts etc.)

Edited May 10 by i_max