Unraid OS version 6.12.0-rc7 available

21 hours ago, EDACerton said:

/boot/config/go

thanks.

6.12 intalled, go file edited with tailscale0 interface added in the included, but can't connect to unraid dashboad with its tailscale's address. Where could I check?

Look under Bug Reports -> Stable Releases

 

There are several topics discussing tailscale

 

 

On 6/7/2023 at 7:37 AM, bonienl said:

 

Ok, found a solution and fix will come in next version.

Thanks

 

I guess its not in 6.12.0 release?

 

I had ZeroTier working fine in 6.11.0, upgraded to 6.12.0.

Through ZeroTier Central I could see that the server was ON, but could not access neither the GUI nor files through SMB, came here and found the tutorial, added the ZeroTier to Interface Extra and then I could SSH and open Unraid GUI through ZeroTier IP but SMB is a no go.
Then I tried with TailScale Docker, same problem, adding to Interface Extra I could access the GUI and SSH but not SMB (even tried using Unraid as an exit node and it worked!).
TailScale Plugin almost the same (could not make it work as an exit node[I did a mistake here, found what I had t do now that I got back to 6.11.0], GUI and SSH ok, but no SMB)

 

I even tried to enable SMB 1.0/CIFS in windows resources, but nothing changed, still no SMB connection.

Reverted back to 6.11.0, restarted the server and SMB is working through ZeroTier and TailScale Plugin (I disabled SMB 1.0/CIFS before testing again in 6.11.0)

 

I skipped 6.11.5 because this Unraid is in production and the next change would be to add a Cache pool, and since it would change in 6.12.0 so I waited.

 

TL/DR:

I use SMB through ZeroTier. Upgraded from 6.11.0 to 6.12.0 followed the tutorial but could not get SMB to work throught ZeroTier. (SMB Local worked fine all the time)

Edited June 17, 2023 by WarpOs
Grammar

I'm really liking the update in general!

Especially the revamped storage configuration is now super intuitive.

 

However the new network interface configuration where unraid doesn't listen on my wg2 interface (which I configure manually on array start for split tunneling) doesn't seem like a good idea in this state.

It would potentially be favorable to allow configuration to listen on all interfaces by default again.

 

I've only had headaches, troubles and why the hell isn't this working moments and it resulted in many pulled out hairs.

Update: I managed to get the reload script working after some fiddlng with how my wg interface is started up.

 

However, a fixed timeout in the reload-script that is configured will at some point produce unreliable results and isn't something I feel confident in.

 

I can see the security improvements that this brings, however for unproblematic operation I would wish that web services are automatically able to start listening to interfaces as they become available that have been configured.

 

I am usually not able to access the server locally, so I'm not quite sure how this is supposed to work out for me in the long run.

 

Update 2:

Nope, still broken. SMB does not work.

 

Update 3:

Now everything is completely broken again and I can't get it to work at all anymore.

Edited June 17, 2023 by Spiritreader
clarification.

10 hours ago, SimonF said:

Are you still on rc7? have you tried the stable release as I think this may have been update in rc8+

I am not. I'm currently running the stable release version. I was just talking to @bonienl on the bug report "[6.12.0] TAILSCALE: SMB, SSH, AND GUI UNABLE TO CONNECT" that I created. bonienl informed me there will be fix for the SMB in the next release. However, I added the code that was provided to go into the 'go' file. See below, this is what my 'go' file looks like:

#!/bin/bash
# Start the Management Utility
# reload services after starting docker with 20 seconds grace period to allow starting up containers
event=/usr/local/emhttp/webGui/event/docker_started
mkdir -p $event
cat <<- 'EOF' >$event/reload_services
echo '/usr/local/emhttp/webGui/scripts/reload_services' | at -M -t $(date +%Y%m%d%H%M.%S -d '+20 sec') 2>/dev/null
EOF
chmod +x $event/reload_services
/usr/local/sbin/emhttp &

This did not work for me when I restarted the array. Docker services and auto start is setup correctly. Did I enter the code correctly?

Edited June 17, 2023 by DC_Interstellar

Thanks for adding the custom-interface option and great to see it made it into the final release. So far 6.12 works great for me after updating from RC

Hi,

Just to make sure I understand (I'm not an expert).

On 6.11.5, you have Tailscale docker running and that's all. So I can connect on my phone for example.

On 6.12.0, you have Tailscale docker running and you have to add something extra interface and edit a config file somewhere (this part looks very complex for me, no clue what is the "go" file) ?

Is that correct ? If so, I'll be sticking with 6.11.5 for a while yet...

Edited July 1, 2023 by zarkoff

6 minutes ago, zarkoff said:

Hi,

Just to make sure I understand (I'm not an expert).

On 6.11.5, you have Tailscale docker running and that's all. So I can connect on my phone for example.

On 6.12.0, you have Tailscale docker running and you have to add something extra interface and edit a config file somewhere (this part looks very complex for me, no clue what is the "go" file) ?

Is that correct ? If so, I'll be sticking with 6.11.5 for a while yet...

You can also use the plugin, which does all of the work for you. 😃

6 hours ago, EDACerton said:

You can also use the plugin, which does all of the work for you. 😃

 

So I install the 2 below and that's it ? no need to edit extra interface and go file ?

 

You have to pick one or the other (they’ll fight if both are installed):

 

- Docker version + extra instructions

- Plug-in version

@EDACerton ah ok, thank you

Sounds like installing Plug-in version is a way easier then.

Thanks to the owner of the repository 🙂

i'm really hoping there will be a fix for the smb issue..

before going to 6.12 (and all rc's above) the smb through zerotier was working perfectly between my 2 servers..

now its not.. i also have to do a manual tailscale set --accept-dns=true everytime i reboot else the dns will not work.. even though i have it set to yes in the tailscale settings..

Just tested zerotier (again) and SMB works fine.

Make sure you have correctly set up the zerotier address under Network Extra.

 

2 hours ago, bonienl said:

Just tested zerotier (again) and SMB works fine.

Make sure you have correctly set up the zerotier address under Network Extra.

 

 

well this is what i got.. and when i run mount i get the error..
 

 

 

 

 

 

Edited July 14, 2023 by furian

Do you have netbios enabled?

Problem with netbios is twofold (1) it doesn't support ipv6 and (2) it wants to use a broadcast address, which doesn't exist on a tailscale tunnel.

 

Netbios needs to be disabled and use WSD instead for host discovery.

 

15 minutes ago, bonienl said:

Do you have netbios enabled?

Problem with netbios is twofold (1) it doesn't support ipv6 and (2) it wants to use a broadcast address, which doesn't exist on a tailscale tunnel.

 

Netbios needs to be disabled and use WSD instead for host discovery.

 

 

 

well... just shoot me lol.. i wasted so many hours trying to figure this out.... i am eternaly gratefully sir..

Hi folks. I followed all the steps here (except editing the .go file because I don't know what file to edit).

 

It says it is listening now on br0, tailscale0

 

However, I am still not able to use samba or login to my Unraid via its IP or tailscale DNS.

 

This is a huge dealbreaker for me. Previously it.just.worked...

 

I tried to get around this and not rely on this but recent events mean that I absolutely NEED this to work with Unraid. What am I missing here?