ptmurphy

Been running for a week or so with no issues have turning of C-States in the BIOS. Weird how this popped up. This server has been running for 3 years without the issue, but with the 6.12.3 and 6.12.4 versions seems to have introduced the issue to my server. Like you said, probably a Linux Kernel issue. Thank you for the help!

I will give this a try and see what happens. This server has been running without issues for about 3 years. Would it be normal for the problem to start so long after a stable configuration?

In the last few days I have noticed Unraid not being responsive after a few hours. On startup everything works fine - shares, dockers, plugins, VMs, etc. However, after 2-3 hours, it goes totally unresponsive - computer is still on, but can't ping the IP address and nothing displayed on the screen. I did some searching and found where the Docker Custom Network Type set to MACVLAN can cause issues, so changed that to IPVLAN, but the problem is still happening. I am running Unraid Version 6.12.4 (updated when problem was happening on 6.12.3). Any other suggestions on what to look for? I have attached the diagnostics in case that will help. tower-diagnostics-20231105-1942.zip

I will keep an eye on it and see if it happens again. Updated to 6.12.3 today, so hopefully that stabilizes things a bit.

Thanks Jonathan. Appreciate the input.

Every once in a while one of my cache drives (in a pool by itself) also shows as an unassigned device. When this happens, nothing on the cache drive is accessible. A reboot fixes it, but any idea what could cause this?

I have reached a point where my VM vdisk sizes are getting too large, taking up almost all of my cache drive, therefore want to move them (or at least the main Windows 11 VM that is taking up most of the space) to another location. I am putting in a 2TB Samsung 980 NVMe M.2 drive and was curious what recommendations this group would have. I have thought of these three options, and not sure what makes the most sense... 1. Upgrade my cache drive to the 2TB, removing the old 1TB SSD cache, and basically run as I am today with VMs on the cache drive (going from 1TB SSD to 2TB NVMe) 2. Add the NVMe drive (leaving current cache for cache and docker only) using the NVMe to store the vdisks. 3. Add the NVMe drive (leaving current cache for cache, smaller VMs, and docker only) converting the Windows vdisk to a physical disk and running it on the NVMe drive. 4. Other thoughts or suggestions? What would you do?

Thanks for the reply. I did finally get it to work. Seems like Unraid didn't like trying to switch from VNC to passing through the video after initially creating the VM. I started from scratch, passed through the video card, and just for safety sake passed through the VBIOS was well.

Does it get all the way through installing? If it hangs at the beginning or in the middle of the installation, I had a similar problem and when installing, choose the "Safe Video" (or something like that) mode.

I successfully upgraded my video card (from Nvidia 970 to 3060 TI) and mostly everything is working (Unraid, Windows VM, Plex). However, I have setup a Ubuntu VM and having a little problem. When using the 3060 TI as the video card, all works well until I install the Nvidia drivers in Ubuntu. Without the Nvidia driver it displays in 4k, etc. I tried to install the Nvidia driver, and upon reboot I just get a black screen. Note that the Nvidia 3060 TI is a secondary card. When I set the VM back to VNC for the video card, the VM will only boot into installation mode, trying to re-install Ubuntu. Any suggestions?

I did the upgrade, it was fairly simple. I took the following steps, and it wasn't too bad. After swapping the video card, I did have a bit of trouble getting my machine to boot from the USB drive initially, but that was another issue. 1. I backed up my VM image file, vm XML setting file, and my Unraid USB stick to be safe 2. Removed Nvidia drivers from the VM 3. Shutdown VM 4. Shut down Server and swap out video card 5. Edit VM setup to pass through RTX 3060 Ti and its corresponding sound card 6. Reinstall Nvidia drivers on VM When you boot into the VM for the first time, you will probably be stuck at a low resolution and get error 43 in device manager (assuming your VM is Windows) until you do step 6 and reinstall your video drivers from Nvidia. That is all it was for me. Everything went well and all is working as expected. Hope that helps...

I had a similar question a few days ago in this thread... I haven't received any replies, so tomorrow I am going to proceed with the upgrade. I will report back here with any problems I run into and pointers based on my experience.

I have finally secured a new video card and am planning to swap out my old video card and want to make sure I do this correct and understand how to do it before I start. I have two cards currently in the system... 1. Nvidia GeForce GT 710 used by Unraid 2. Nvidia GeForce GTX 970 - passed through to VM, not using VBIOS dump since it is a second card. I am using PCIe ACS override (set to "Both") to pass through an NVMe drive, a SATA controller (for Blu Ray drive), and a USB Controller for mouse/keyboard, and any other attached device such as an external drive. I am replacing the GTX 970 with an Nvidia RTX 3060 Ti. Here is what I was planning to do... 1. Remove Nvidia drivers from the VM 2. Shut down Server and swap out video card 3. Edit VM setup to pass through RTX 3060 Ti 4. Reinstall Nvidia drivers on VM Are there any steps I am missing so that Unraid forgets the old card, sees the new card, and allows it to pass through the VMs?

Thank you for the information. I hooked the drive back up in the usb enclosure and it still worked, so seems the 3 pin 3.3v power was my issue. I did cover pin 3 (more accurately pins 1-3 - dang those pins are small. Couldn't get the 1-pin wide tape to stick, but covering 1-3 worked), plugged the drive in, and it worked! Again, appreciate the help!

I adding a fourth 8TB drive (1 parity, 3 data) to my array. All 3 previous drives have been shucked WD EasyStore drives and went into the array with no problems. This fourth drive (same 8TB shucked WD EasyStore) works on USB, as well as an external eSATA dock, but when connected to an internal SATA power cable, it does not power up (have tried several cables that other working drives are attached to, and still no power to that drive). In my research, it appears that there is a new standard on some drives that if 3.3 volts are applied the drive does not turn on, as referenced in this thread... However, I can't find how to actually fix this. It appears that the third pin on the SATA cable/connection to the drive needs to be disabled. Is that correct? If so, what's the best way to do that without any permanent damage? Also, how do I verify that this is actually the issue?

Just to add my experience in case someone else is still having troubles... I upgraded to 6.3.2 today and had similar problems to others in the past. I ran the script and the USB stick would not boot. Said the Linux Kernel was missing. Here are the steps I had to take to fix it... 1. I took the USB stick out and put it into a Windows Machine. 2. Downloaded the latest version of unRAID from the website (6.3.2) 3. Tried to copoy copy the syslinux folder from the download to the USB stick, which failed 4. The USB stick had what appeared to be a file called syslinux on it already, so it would not copy the syslinux directory to the USB stick, so deleted that file on the USB stick 5. Successfully copied the downloaded syslinux directory to the USB stick 6. Ran make_bootable.bat from the root of the USB stick 7. Let it compete and safely "remove hardware" and eject the USB stick. 8. Put the USB stick back into the unRAID server and booted up. From that point on, everything worked fine. Thanks for the help from the previous posters in this thread.