bastl

Fix confirmed 🥰

@Fma965 With Unraid Update 6.9.0 and next point release 6.9.1 Your plugin stopped working for me. Nothing changed. I use a H1200i and never had an issue before showing the stats on the dashboard. Any idea how to troubleshoot and to fix it?

double Thanks for the temporary fix 😂

Thanks. Does this survive a server reboot?

I've noticed the same thing. Syslog in the WebUI starts with some entries with plugin update checks. Would be nice to have the full syslog since boot back in the WebUI. I guess this was set in the RC builds where people reported spams in the logs to avoid the webinterface breaking.

Thx for the info. Is there any settings to show again the full syslog inside the WebUI? Edit: I was to quick. Noticed your edit to late. Thanks for the link

Update from 6.8.3 went smooth as always. Temp sensors on TRX40 are working now. 😘 Switching from old pci-vfio in syslinux config to the new method also went without any issues. But a small thing I noticed. Syslog isn't starting with the normal boot process entries. The first entries for me are from some plugin checks/updates Is this normal behaviour? Tried it with Firefox (main Browser) and Brave (never used for my server before). Same result. Is there a max line limit now for showing the logs and I can't find the settings?

@Chenhi Be careful with this. Don't blindly passthrough what is shown. If the GPU for example is in the same group with one of your NVME drives or an storage controller this will cause your server to crash and in the worste case a data loss. Only passthrough devices which really belong to the device which you want for example the USB controller and the audio part of your card. Sometimes it might help to use another PCIe slot for the device you wanna passthrough.

@Jaster I bet you copied the xml of a running VM and used it for other VMs. Default entry for a none running VM for VNC looks something like the following: <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0' keymap='de'> <listen type='address' address='0.0.0.0'/> </graphics> The "port" option is set to auto and counts up with each new VM you start. If you hardcode a value in, it tries to use that port and if in use by another VM you get that error. If I look into a the same xml from that VM while online it looks like the following: <graphics type='vnc' port='5903' autoport='yes' websocket='5703' listen='0.0.0.0' keymap='de'> <listen type='address' address='0.0.0.0'/> </graphics> This is the 4th VM on my system which uses VNC and therefore it got port 5903 starting with 5900. If you need some fixed ports, try to use higher values depending an the amount of VMs you're using to prenvent conflicts and set the rest of the VMs to auto like it's shown on the first screenshot. Make sure your keymap is for your language.

@GoustiFruit Check your AdBlock Lists (I guess pihole). For me there was an Microsoft Server blocked used by unraid to check network connectivity and prevented me to update Unraid. I had to whitelist "www.msftncsi.com". Maybe the process to activate Unraid uses the same server for checks.

@GoustiFruit Make sure your server has internet access. Trial versions don't run without an active internet connection.

@frankie666 Maybe try a fresh new Windows iso. In the last couple years I had 2 or 3 directly downloaded from MS which won't be able to install. Use the Media Creation Tool from Microsoft to download the current build. Maybe this helps. Btw, how large is your cache drive and how much space is free?

@frankie666 Make sure when you first create a VM to define the vdisk size in GB. If you don't enter a value a vdisk is created but only a couple MB. Also make sure not to use core0 for the VM. This core is always used by Unraid itself to do stuff in the background and can end in low performance for the VM. With the following command you can show more infos about the created vdisk. Make sure the virtual size matches the value you defined and doesn't show a couple megabytes only. qemu-img info /mnt/cache/VMs/Mint/vdisk1.qcow2 Adjust the path for the vdisk. In my example it's an qcow2 disk, but doesn't matter.

As long as you use VNC + GPU passthrough you won't get any video output from your GPU. Remove the VNC GPU and report back.

@Schmeckles23 Unraid 6.9 will support the sensors for the newer Ryzen chips. You can already use one of the BETA builds, wait for the stable version or fiddle around with a custom kernel. It's up on you 😄

Small observations I noticed for a couple weeks now. Every morning at 6am I do a automated BTRFS snapshot for my VMs on the cache drive to an UD drive. This works for over a year without any issues so far. What I have noticed starting on 23.11.2020 is that the allocation of the UD drive isn't shown correctly when I visit the main page after the snapshot anymore. The snapshots are created as usual (70-80G per day) but the allocation isn't showing the new values. I see the same values as the day before. Refreshing the main page isn't triggering a rescan of the drive. I am on 6.8.3 and installing plugin updates as they come in, nothing else has changed. Almost every day I check for updates. If I do a "BTRFS sub list /mnt/disks/backupdrive" for the UD drive the values are shown correct again. If I don't do this, the old values are still shown. I never had to do this before. For me this is a small problem, because I have to clean old snapshots once every 1-2 months by hand on the 1,5TB drive to not fully filling it up. @dlandon I know it's not really a big think, but maybe someone can explain me whats the reason for this new behaviour.

What devices are you passing through to the VM?

@doesntaffect What devices are you passing through to the VM? Any USB devices, controllers or disks? A complete diagnostic would be helpful

Remove that part from the "cpu mode" section from your xml so it looks like the following and try again with more cores. <topology sockets='1' cores='1' threads='2'/> Not sure how you got that option in there. I never saw Unraid adding that option by itself.

@Fabiolander Do you have any VMs running? Can you try to shut them down and restart the docker engine. I saw a comment yesterday a guy reporting that some WebUI elements in unraid not showing up or for example only 2 out of 3 running VMs are showing on the Dashboard. Maybe you have some similar issues caused by the VM engine.

@Fabiolander Maybe try the "binhex krusader" docker from the app store. Looks like you using the older version. I use the binhex version for quite some time now without any issues. Not sure if the old one is still maintained but I think this was the reason I switched to the binhex container.

I know it's using NoVNC. Sry, my bad, maybe the wrong wording. I wanted to say there is no so called "NoVNC" option for me.

@Fabiolander I never came across a docker where a VNC option is available. Krusader for me if started shows a WebUI link which perfectly works fine to access the docker.

@abhi.ko As long as the OS isn't fully loaded no tools like Splashtop or Teamviewer will work. Easiest way is to passthrough a USB keyboard (wireless won't work).

@cracksilver Small hint, don't give the VM all the cores your server has. Core 0 is always used by unraid itself. You better uncheck 0/8 to reduce interferences with backround tasks running on the host system.