ClintSouthwood

Hello fellow unRAIDers, I'm humbly asking for help from the gurus here. I've spent several days trying to solve this one but have come up empty. Any help would be greatly appreciated! I've been using a few rClone mounts for around a year, happily serving up files from gDrive to Plex and to a SMB share without issue. A couple days ago I decided to finally set up Intel AMT in the server's BIOS to allow me to control it over the LAN. Upon rebooting, my trusty rClone mounts were no longer accessible over SMB or from within Plex. Windows now complains about permissions and can't access the shares. Plex also cannot see the shares - when attempting to add a new library, the shares are shown but they are greyed-out and have a document icon next to them. Research on the web about the greyed-out shares turns up a freenas user who had the same issue and fixed it by changing the permissions of the Samba share from within Windows. This obviously won't work for me as Windows can't access it. My rClone mounts were located under user shares. Now, when I first set these up I didn't know about the issues with mounting shares under /mnt/user but after reading a lot and trying to solve my issue I now know that this is bad form and many people advise against it. I think that I understand the potential issues that can arise from mounting in /mnt/user but I fail to see how it could adversely affect me in this situation. Could this somehow be the root of my permission issues? I've since changed the mount target to be under /mnt/disks but have had the same issues no matter where it is located. I've also updated my rClone plugin and discovered that there are many new options since I originally started using it, including ways to change ownership and permissions using the --uid, --guid, and --umask flags. I've tried to use these in my mounts to make them appear to be owned by nobody/users with max perms but Windows still fails complaining about perms and Plex still shows the shares as greyed-out. The console reflects the ownership/perm changes from these flags when doing a directory listing, so it does appear that the flags are being applied. I've also tried using chown and chmod to change the ownership and perms of the parent directories from root:root to nobody:users without success. The funny thing is that shares and the files within them can be listed and manipulated via the console (as root) without issue. Furthermore, Plex can see the shares ONLY if the docker is run as root (uid=0 guid=0 umask=000) and Windows can access the share only if force user = root and force group = root exist in samba conf. These clues lead me to believe that the rClone flags are just not working properly even though a ls -l in the console shows the mounts owned by nobody/users with full perms. I'm out of ideas on how to fix this issue. It's very odd because there was no clear cause for this issue. My last ditch effort is to try and restore my appdata and even perhaps flash data from a backup created before this issue occurred but I have doubts that will do much. And, I've never used the CA backup/restore plugin so I'm a bit reluctant to try this before asking for help. I can provide additional info such as rClone mount flags and/or SMB params but none are particularly exotic and this was all working flawlessly for many months so I doubt it's a simple config issue. I haven't been able to find any reports on the web of similar issues where rClone fails to properly spoof these parameters so I'm wondering if it may be unRAID specific. Thanks in advance!

You can add the mounts under settings - SMB - Samba extra configuration or by editing the conf file directly. Spaceinvader has a great video on it. I don't think you can reliably write to the shares though. I somehow broke mine recently and can only access my rclone mounts via the terminal.

Would it be possible to add sas2flash for HBAs?

Super stoked to see this issue get some attention. I too use SAS disks and would love the ability to spin them down. I tried the start/stop commands above on a 6.8.3 installation. The spin down command seemed to work as smartctl reported that the disk was not ready afterwards. Unfortunately, the disk in question refused come back up using the spin up command and was subsequently disabled by unraid. I suppose it's possible that I didn't wait long after issuing the command though. I'm currently doing a read check to try and re-enable the disabled disk.

Worked like a charm. Many thanks!

When attempting to run iperf3 I'm getting the following error: "error while loading shared libraries: libssl.so.1..." How can I add the requisite libraries? TIA

Thanks, I will check these out - they both look promising. I didn't consider using a separate container for monitoring though it makes sense. The support thread for the Munin-server container by Jugniji in CA seems to have not been updated for quite a while but there are several other templates available in dockerhub. I figured that displaying temperature readings would be a natural extension of the system stats plugin and I was somewhat surprised when I couldn't find any mention of it. Since I have both plugins currently installed and configured, I thought perhaps there was simply a configuration setting I missed to integrate the two for unified data on one display. The idea being that one could find a peak in CPU usage, say during an intensive transcode operation, and correlate the peak usage with the peak temp.

Thanks for the info, will look into this as I thought pads were generally inferior to TIM compounds. Now if only I could find a better heatsink for my mobile chip! My question still stands though, would love to have historical temp data available in some form.

For what it's worth, I noticed the same thing (the path was not populated by default) but it does exist in my container.

I currently have both Dynamix System Stats and System Temp plugins installed and they work great. Many thanks. Apologies if this has already been asked but my search didn't turn up anything. Is there any possibility to integrate historical system temperature data into the System Stats graphs? I would love to be able to monitor temperatures over time. If not, can anyone suggest something else that can enable this functionality? Thanks in advance.

Fair enough, I do see it in the post now, thanks for linking it. I did skim this post before upgrading but obviously missed that pertinent bit. I'll certainly be more careful going forward. Still, removing current (core/key) features, no matter how "improved" they are, is pretty low, IMO. This is not new behavior from the Emby team and, quite frankly, I'm not surprised - they've been on this trajectory for quite some time now. I respect developers and understand that they want to be paid for their effort but to change the entire business model just after gaining a decent userbase (most of whom likely jumped ship from Plex for similar reasons) is disturbing. I sincerely hope that Jellyfin takes off, it's mere existence indicates that I'm not the only one upset with the transition of FOSS into a closed-source, subscription-based model. Apologies for the rant, I know this isn't the place for it.

Thank you for all your awesome builds, kind sir. Will migrate over to your container ASAP.

Mostly a lurker here but I didn't see this mentioned yet. The latest update seems to have removed the ability to use hardware transcoding for non-premium users. I really dislike the underhanded nature of placing current features behind a paywall and neglecting to mention their removal in the changelog. I've begun testing the Jellyfin fork though I had to set the docker up manually since there isn't a prebuilt one in the community apps yet.