tjb_altf4

Not using this docker so I can't test, but this should get you setup. Looks like you need to create an appdata config path like most dockers in unraid are setup AND create a variable with the key of SYNAPSE_CONFIG_PATH pointing to the container side path i.e. /config as the value If postgres is installed you need to create the synapse database user and database itself. You will need to use the console for the docker container to complete this OR you can use the GUI tool like pgAdmin Documentation details what needs to be done: https://github.com/matrix-org/synapse/blob/master/docs/postgres.rst It also details porting from sqlite, but I would start clean slate. Dockerhub details the container variables you need to pass for postgres, the snippet below should be enough to get the synapse talking to postgres. https://hub.docker.com/r/matrixdotorg/synapse/

This is how mine is setup and it works flawlessly, in my opinion it takes a lot of IO stress away from the cache & array. note: I'm using Sickchill/CouchPotato, but same principal applies.

@saarg I nuked both the image and the data and started fresh with latest. Looks like postgres needs to be brought up to v11. 2019-03-05 07:49:20.931 AWST [79942] FATAL: database files are incompatible with server 2019-03-05 07:49:20.931 AWST [79942] DETAIL: The data directory was initialized by PostgreSQL version 11, which is not compatible with this version 10.5.

I like the immutable attribute solution, but a bandaid is to pass media shares to the docker as read-only instead of r/w. This will have some limitations, but is a quick solution to implement to secure data.

Same as above scenarios except my game rig is a physical machine (for now). 90% of my library sits on the array, the rest lives on my local install. This usually happens for performance reasons, or in other cases some games simply don't work (or install) without block level access to drives.

While nested virtualization is supported, it's no longer turned on by default, however there's a user script to enable/disable as required.

If I refresh after selecting new icon it goes back to the previous icon, the XML is not being updated at any point.

Should be able to extract via backend using a sql query, or import/export via gui sqlite tools.

Which plugins? The maximum in the plg file might have been set to 6.6.6 Only Dynamix System Stats now, other one updated for me just then.

Only issue I've come across is a few plugins that have been updated recently won't allow install on 6.6.7 plugin: installed Unraid version is too low, require at least version 6.7.0-rc1 Not a biggie, but something to be aware of.

Yep I got it up and running last year, I dare say the db setup or seed is the failing point. The guide linked earlier covers everything I did to get it working.

Take a look at jellyfin, its a project forked from emby 3.5.2. If your not happy with the direction emby has taken, this is a good alternative.

+1 on Android icon, was just about to post to see if nuking my app was warranted. iOS icon still correct.

New image seems to be working as intended and can see appdata folder being populated as expected, thanks!

@binhex I'm seeing the same issue mentioned above: Jellyfin appdata template Jellyfin appdata For comparisons sake, here is Emby appdata Run command (formatted) root@localhost:# /usr/local/emhttp/plugins/dynamix.docker.manager/scripts/docker run -d --name='binhex-jellyfin' --net='br0' --ip='192.168.1.212' -e TZ="Australia/Perth" -e HOST_OS="Unraid" -e 'TCP_PORT_8096'='8096' -e 'UMASK'='000' -e 'PUID'='99' -e 'PGID'='100' -v '/mnt/disks/downloads/_temp/emby-test/media/':'/media':'rw' -v '/mnt/disks/downloads/_temp/emby-test/tv/':'/media-tv':'rw,slave' -v '/mnt/disks/downloads/_temp/emby-test/movies/':'/media-movies':'rw,slave' -v '/mnt/disks/downloads/_temp/emby-test/music/':'/media-music':'rw,slave' -v '/mnt/user/appdata/binhex-jellyfin':'/config':'rw' 'binhex/arch-jellyfin' c727eb966e562c0f8c20a88bbdc36eada63cd79edbdf53b9d29303c9c9602138

@limetech could this be leveraged to change the default GPU (and presenting another) for the unraid GUI console?

As this is external storage and nextcloud is accessing via SMB over network, the share must be accessible on the network (as set in Share settings). Internal paths won't work, and I would personally create a non-root user to access those shares. Assuming the share is exported to network, it should look like /media/ not /mnt/media/

Sometime browser autofill gets a little over zealous, I was filling in the letsencrypt docker with my email and it autofilled the docker name field with my name... so I ended up with a docker with my name 🤣, I did have a good giggle when I worked it out. Might have been something similar in this case.

I just had a quick read around and it's inclusion is aimed at VR headsets for power/video/data instead of a birds nest of cables, dubbed VirtualLink. Being aimed at VR probably explains why it's only on the higher SKUs. https://hothardware.com/news/virtuallink-vr-headset-usb-c-connector-spec-backed-amd-nvidia-microsoft-oculus Pretty exciting bonus for virtualisation users, hopefully AMD picks up the standard for it's cards to!

RE: using delugevpn privoxy with jackett On the LSIO docker I did this way, @binhex might be able to comment whether something similar can be done or added to this container. I remember having issues with the in-app proxy settings, can't remember what though

Should work, I used one of the HyperX PCIe M.2 SSDs in my previous unraid build (x58 board), which has this kind of adaptor.

Nextcloud (docker) might be a nice way to handle this, particularly for the less technology literate users you may want to share content with.

I've had gskill (samsung b-die) ram running at it's vendor rated speed of 3200mhz @C14 on 1st gen threadripper unraid build for 11 months. Has run without issue and was tested heavily before going "live" on my system. My 2c, don't cheap out on ram, particularly if you want to run above base supported speeds. Definitely would recommend running extended memtests once you make your purchase to validate all is ok, I ran mine for nearly a few days.