mackid1993

That's strange, I personally haven't had any issues although I'm running the beta.

There is a known memory leak with drivers below 248.

Update your virtio drivers.

Ah so the key is <model fallback='allow'>Skylake-Client-noTSX-IBRS</model> does that impact performance?

Do you have an Intel CPU? I've been trying to run WSL on a 12700k for so long now. Is this working now in 7.0.0 beta2?

It's not possible with Windows. The second you enable hyper-v the VM doesn't boot. You have to go in to recovery and run dism /Disable-Feature /FeatureName:Microsoft-Hyper-V /Image:C:\ at a command prompt.

Not sure, maybe try the Unraid beta?

Yup that's it. If you are seeing CPU core spikes and gradual sluggishness with your Windows VM and other VMs you'll want to look into hugepages.

Have you tried on Unraid 7.0-beta1? There are a lot of improvements @SimonF made. If you are still having stability issues try my guide to enable hugepages, you just need enough RAM to always reserve and have to mind any docker containers (usually anything with a database engine) that will use them before the VM can.

Fair enough. Looking forward to an official implementation!

Hugepages also fixes memory fragmentation that causes other VMs not in the shared memory pool to get sluggish.

Just a shot in the dark, maybe try hugepages to see if it helps you.

I would open a report with the Virtio-Win team on Github. Clearly there is a bug with AMD hardware. I would help but i'm an Intel guy.

Another user @johnsanc was having major speed issues on AMD. I wonder if there is some sort of issue with the virtio-win driver and AMD platforms.

It may be a cache issue. Look into this: and @SimonF's php script and bash script that will be added in 6.13/Unraid 7. It works in 6.12 but configuration is a little advanced. You'll also need to download rust virtiofsd which will also be in 6.13/Unraid 7. If you look through past posts there's a lot of information to configure this in 6.12.

Honestly it's much easier and safer for your Apple ID to go on eBay and buy an old Mac Mini or Macbook Air and use that to host BlueBubbles. OCLP is great for installing a newer version, currently the best version to use for BlueBubbles is Ventura and it's pretty much set and forget. I keep a Macbook Air with a dummy HDMI plug and ethernet adapter sitting on top of my Unraid server with an iPhone plugged into it charging for phone number registration. My setup is bridged into Beeper and everything works great.

No problem. I had an extra few minutes today. 😊

@wacko37 1. Stop the VM and Docker service. 2. Go to Settings -> Network Settings 3. Click add VLAN 4. Fill it out and click apply. I already have mine set up so my VLAN number for this example is 3, yours will be 2. 5. Change iP address assignment to static: 6. Enter 10.0.0.1 this will be the address of your Unraid server on the VLAN. Click apply 7. Scroll down and find your gateway, likely br0.2 8. Start VM and Docker back up. 9. Edit your VM template. Scroll to the bottom and add a second NIC. Set it to virtio and the second network source to your gateway likely br0.2. 10. Start your VM, open a run dialog in Windows and type ncpa.cpl. Find your new NIC that doesn't have an IP address and give it one as follows. Just check IP addresses to make sure you don't set a static address on your main NIC. Now your server can be accessed over the VLAN from \\10.0.0.1. However when a lot of activity is happening since SMB is single threaded other applications will slow down. To work around this, I create hosts file entries for each application I use and authenticate separately for each one. Open an admin command prompt and type: notepad c:\windows\system32\drivers\etc\hosts These are my entries but you can make as many as you like. Once this is done in Macrium you connect using \\macrium\share and then your Macrium traffic will stay completely within your server and you can avoid any compatibility issues related to Virtiofs and instead use Virtiofs in situations where it works best.

Probably not a major difference, it will work if your router doesn't. That's about it.

You don't need a second NIC. It's a virtual NIC that you need. I'm sorry I'm not sure why you are having issues.

I usually use SMB within Macrium for backups. I suggest creating a vlan on your server and setting a secondary virtual NIC to it. Then setting a manual IP address in Windows and creating a hosts file entry for Macrium that points to the server using that vlan. That way your SMB traffic doesn't leave the server.

I thought I'd share this here. I wrote a pretty easy to use mount script for rclone that will both mount a remote and start any dockers. It's also designed to be run on an interval such as every hour to ensure that the mount is still active and that the dockers are still running. Here it is if anyone finds it helpful: #!/bin/bash # Define the mount directory MOUNT_DIR="/mnt/remotes/mount" # Define rclone remote name RCLONE_REMOTE="remote_name_here" # Define rclone cache directory location RCLONE_CACHE_DIR="/mnt/cache/appdata/rclonecache" # Define Docker containers as an array. Containers should be in quotes seperated by spaces. DOCKER_CONTAINERS=("docker1" "docker2" "docker3") # Check if the mount directory exists, create it if not if [ ! -d "$MOUNT_DIR" ]; then echo "Mount directory does not exist, creating it..." mkdir -p "$MOUNT_DIR" fi # Check if rclone is already mounted at MOUNT_DIR if mount | grep -q "$MOUNT_DIR"; then echo "rclone is already mounted at $MOUNT_DIR." else # Mounting the drive echo "mounting rclone at $MOUNT_DIR." rclone mount \ --allow-other \ --allow-non-empty \ --log-level INFO \ --poll-interval 1s \ --dir-cache-time 1m \ --cache-dir="$RCLONE_CACHE_DIR" \ --vfs-cache-mode full \ --vfs-cache-max-size 100G \ --vfs-cache-max-age 24h \ --vfs-read-chunk-size 128M \ --vfs-read-chunk-size-limit 1G \ --vfs-read-ahead 128M \ --uid 99 \ --gid 100 \ --umask 002 \ $RCLONE_REMOTE: "$MOUNT_DIR" & fi # Ensuring the mount operation has enough time to initialize before starting the docker containers sleep 10 # Loop through the Docker containers array for container in "${DOCKER_CONTAINERS[@]}"; do # Check if the Docker container is already running if docker ps | grep -q $container; then echo "Docker container '$container' is already running." else echo "Starting Docker container '$container'..." docker start $container fi done Unmount script (to be run at array stop): #!/bin/bash # Define the mount directory MOUNT_DIR="/mnt/remotes/mount" # Unmount the mount point fusermount -uz "${MOUNT_DIR}"

Trial and error. Allocate enough just for the VM, and start containers one by one then try to start the VM. Which ever ones make the VM not run are using hugepages. If you can post them here let us know, I'll update my post.

That should help you. It makes a big difference and is quite easy to set up. Just make sure you have enough memory to reserve.