Iker

Hi, that is an issue with the x470 board BIOS, in my older mobo I had the exact same Processor and board, a x470 Gaming Pro Carbon with 2.E0 Bios; you have to enable the option "AMD Cool & Quiet" ; "PBO" and "Global C-States" must be Enabled; make sure that 'Power supply idle mode' is set to 'Typical Idle Current'; then the ACPI Driver could work nice and let you choose the profile that you prefer. Even if I haven't checked your ram in the QVL list for the MSI Board, as a safe measure, disable DOCP and let the ram at 2133 Mhz, then, when everything is stable, start your way up in the ram speed, as an advice, you should never go beyond 3200; 3600 Mhz is way to high for the Processor and Unraid in general (I could be wrong, but there is not much performance left in the table beyond 3200 with a 3900x). Let me know if everything works ok.

Guacamole is a good option, however, you could try using a TCP Proxy - HAProxy will be a perfect solution as well.

I have a Ryzen 3900x, recently I have to create a Windows 7 x32 Machine, and it works without problems.

Hi, i'm also using portainer, the compatibility with a proper compose file isn't complete so probabily you should use docker compose, It is possible to install docker-compose in unraid so you could just download the yml files and "docker-compose up" your docker.

I have a 3900x and the same issue, the only thing that worked was to emulate to Skylake so AES NI is present: <cpu mode='custom' match='exact' check='full'> <model fallback='forbid'>Skylake-Client</model> <topology sockets='1' dies='1' cores='2' threads='1'/> <feature policy='require' name='hypervisor'/> <feature policy='disable' name='pcid'/> <feature policy='disable' name='hle'/> <feature policy='disable' name='erms'/> <feature policy='disable' name='invpcid'/> <feature policy='disable' name='rtm'/> <feature policy='disable' name='mpx'/> <feature policy='disable' name='spec-ctrl'/> </cpu>

Upgrade from beta 35 to RC2 without any problems MOBO: OS: UNRAID OS Basic Version: 6.9.0-rc2 (Next) (on Kingston Data Traveler 16GB) Hardware: X470 GAMING PRO CARBON (MS-7B78) | AMD Ryzen 5 2600 OC 3.8 GHz | Corsair Vengeance LPX (2x32GB) DDR4 DRAM 3600MHz C18 | Geforce GTX 1060 3 GB, Intel 82576 Dual NIC Storage: Intel 660P NVME 1TB x 2 RAID-0 Cache | Seagate Barracuda 4 TB, 8 TB | Seagate IronWolf 4 TB Dockers: letsencrypt, Plex, Tautulli, RedNode, MQTT, Wekan, Pydio, SageMath, Psono, ELK Stack, Sonarr, Radarr, Bazzar, TIG Stack VMs: Windows 10, Windows Server 2019, MacOS Catalina, Ubuntu 20.0.4, pfSense

I can confirm that in my mobo, upgrading from 6.8.3 everything is working flawlessly, the cpu governor and performance in general is better than ever and bluestacks is finally working in my VMs. Hardware: X470 GAMING PRO CARBON (MS-7B78) | AMD Ryzen 9 3900x NO OC | Corsair Vengeance RGB Pro (4x16GB) DDR4 DRAM 2400MHz C16 | Geforce GTX 1060 3 GB, Intel 82576 Dual NIC Storage: Intel 660P NVME 1TB x 2 RAID-0 Cache | Seagate Barracuda 4 TB, 8 TB | Seagate IronWolf 4 TB Dockers: swag, Plex, Tautulli, NodeRed, MQTT, Wekan, Pydio, SageMath, Psono, ELK Stack, Sonarr, Radarr, Bazzar, TIG Stack VMs: Windows 10, Windows Server 2019, MacOS Catalina, Ubuntu 20.0.4, pfSense

Check to start Unraid without the passthorugh in the vfio (Remove completly the video card and audio), and just try to start the VM, it have to take it away from unraid and boot without problems.

Taking a quick look into your diagnostic files, there seems to be a problem mounting the libvirt image cause you don't have enough free space in your cache drive, there are a lot of "Sep 5 20:01:16 Tower root: mount error\nSep 5 20:01:16 Tower shfs: share cache full", probably a good idea to disable the mover temporary, set the Min Free Space setting for the cache a little above and run a balance.

For me it was very straighforward with no need to create records in the DC, just set the DNS to the domain controller, use an AD admin to join it and voila it works!, to clarify the domain controller is running Windows Server 2012 R2 and the schema level if for the same OS.

Be aware that even if the SATA Expansion card reports that the Disks are connected to a SATA 3 Port, they won't be runing at 6 Gb/s, PCIE 2.0 is going to heavily limit the speed of your SSDs and add some latency, for HDDs is okay, try to get a x4 SATA Expansion card if you like to get most of your disks.

I have implemented in a production server for a small company, as long as the server don't restart due to an error or some hardware crash, it persist across reboots and works as expected; but... if the server crash, you have to stop the array and join it to the domain again.

You could try with "DiskSpeed" from CommunityApp or dd command (https://www.cyberciti.biz/faq/howto-linux-unix-test-disk-performance-with-dd-command/)

Thats very weird, I have 2 intel 660p in a BTRFS cache pool in Raid 0 and i'm getting +-2.4 GB/s, what is the raid mode of the btrfs?, how are you testing the speeds?, you could try with the community apps "DiskSpeed", directly with dd command or even a VM to check you disk speed over the /mnt/cache.

No problem, the command line is: xfreerdp /f /sec:nla /u:user /p:password /rfx /gfx /gfx-h264 /multitransport /network:auto -bitmap-cache -glyph-cache /gdi:hw -fonts /usb:auto /v:ipaddress Keep in mind that I set a couple things more: I switch to the legacy driver on the pi4 as susggested in the reference post (sudo raspi-config -> Advanced Options -> GL Driver -> Legacy). The pi4 is connected to my wired network, mainly because is more convenient for me, i haven't tested the setup in Wifi 2.4 o 5 Ghz Try with the wired network, maybe you could solve any latency or choppy video with a wifi stick. Initially i tried wtware and thinlinx but i got mixed results and the video experience wasnt that good;

Unfortunately no, i ran into the same problem, but the drivers included in the kernel do not allow you to control the fan, it requires another tool that although is present it cannot run properly because of the lack of and x server.

I think that this could be accomplish with a user script, specifically with the command "virsh list" and "docker stop", you only have to check is the vm is present or not in the list, the main problem is the delay (VMs in SSDs starts veryfast) of the script, but probably you could just let it in a constant loop, so every 5 seconds checks if the vm is running or not.

Yes, is that simple, i've using NVME disks in my unraid and it detects them without problems as unassigned devices, the only caveats is the support for passthrough to a VM, specifically with Intel 660P and some other models.

I had that exact setup that you suggest for my girlfriend and works just great: Win 10 VM: 8 GB RAM 200 GB Primary Disk (Cache/SSD) - 200 Secondary disk (Array/HDD) 4 Cores - Ryzen 2600 Memory video increased to 32 MB just to be Safe. RPI4 2 GB RAM: Raspberry Pi OS - latest version RDP client - freerdp The video and audio are great and generally the experience is quite smooth. The only trouble that i got, was the command line for freerdp with the audio redirection, if you select the wrong device as output (In the raspberry desktop) the rdp session crash, so make sure that you select the correct one (Bluetooth, analog or HDMI); this post help me a lot with the command line for the freerdp (https://www.raspberrypi.org/forums/viewtopic.php?t=249628).

I have a lot of containers that aren't present in the CA Application, the easy way that i found to install them, was using Portainer, its pretty easy to configure and let you personalize most of the parameters of the container, even docker compose if you like.

Just installed this plugin, it works amazing!, thanks for your hard work.

Upgraded from 6.8.2 without any problems, all Dockers and VMs working fine, also tested the Nvidia plugin and works flawless.