I know this is an older topic, but tried to follow the instructions and guidance given and am unable to get this working.  Perhaps it's a change in unRAID that is preventing this process from working but I wanted to try and see if I could get this to work.  I have update my "go" and "smb-extra.conf" files as noted below.  The share does show up, but is not accessible.  Let me know what I'm missing, please.

 

go file contents:

#!/bin/bash

# Start the Management Utility
/usr/local/sbin/emhttp &

# force iptable mangle module to load (required for *vpn dockers)
/sbin/modprobe iptable_mangle

# Create transcodes share
mkdir -p /mnt/transcodes

 

smb-extra.conf file contents (user has been modified to protect the innocent, using actual username from configured user on server):

[transcodes]
path=/tmp/transcodes
	comment = 
	browseable = yes
	valid users = <my_user>
	write list = <my_user>
	public = yes
	writeable = yes
	vfs object = 

 

The directory is created and I noticed that it is owned by root, so I "chown nobody:users /mnt/transcodes" and "chmod 777 /mnt/transcodes", but that had no change.  Share is still inaccessible from my PC.  My goal here is to present a ramdisk share that is accessible from multiple nodes on my network for tdarr transcoding as opposed to wearing out my temp_pool SSDs that I had setup for this.  If I were only transcoding on my server then I would just pass the /temp path into /dev/shm but that isn't accessible from my other nodes and the jobs would fail.  

You posted 2 different paths for the folder creation and samba config.

Also if nodes are on other machines then they should transcode to their own ramdisks instead of going through the overhead of a samba share.

Edited December 16, 20232 yr by Kilrah

2 hours ago, Kilrah said:

You posted 2 different paths for the folder creation and samba config.

Also if nodes are on other machines then they should transcode to their own ramdisks instead of going through the overhead of a samba share.

I tried running the nodes on their own /temp directory local to them and all jobs on those nodes would fail to copy.  Maybe then my issue is with tdarr?