hovee

I had not rebooted. I just rebooted and everything now works perfectly. I am able to access the Unraid UI from the Tailscale IP while connected from my phone on cellular. Thank you again!

@EDACertonI'm having the same problem where I'm unable to access my Unraid Gui or ping it from a remote device. Other devices on my tailscale account I can ping and SSH in to. It's just the unraid device that I can't access remotely. I do have Unraid Services listen on Tailscale IP set to yes and I see tailscale1 listed on extra listen interfaces on the network settings page. Any thoughts? Amazing plugin! Super appreciate your work! Attached are logs. Unraid-tailscale-diag-20240203-145724 2.zip

I put in a different GTX 1060 and it was recognized instantly in the Nvidia Plugin. It was definitely a bad card. Thank you for your help!

I tried it in another PC that has Windows 10 installed. The card doesn't work in that machine either. No video output. I think the card may be bad. I'll see if I can get it replaced and try again with a new card.

I believe the card is supported as I've read posts and videos of people using a 1060 and a 1080 in my server HP z620. I upgraded the bios to the latest version today. I didn't see any options for 4g decoding or resizable bar in the bios. I did try changing a setting for 'Video Options Rom' from Legacy to EFI. That didn't seem to make a difference. The power supply is 800watts and should supply enough power. An older quadro card I had works in the pci slot and displays video output of unraid booting, etc..., so I know the pci slot is ok. I'm going to try the 1060 card in another PC I have to see if it works over there as I'm wondering if the card is bad.

I also tried installing driver version 470 and unraid crashes as soon as I click on the plugin in settings.

My Nvidia GTX 1060 card is not showing up in the Driver Plugin. I can see the card in System Devices, but in the plugin it says no devices were found. I do however see this error in the logs: Dec 21 14:11:15 Unraid kernel: nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint. Dec 21 14:11:15 Unraid kernel: nvidia-uvm: Loaded the UVM driver, major device number 239. Dec 21 14:11:16 Unraid kernel: NVRM: GPU 0000:05:00.0: RmInitAdapter failed! (0x24:0x72:1436) Dec 21 14:11:16 Unraid kernel: NVRM: GPU 0000:05:00.0: rm_init_adapter failed, device minor number 0 Dec 21 14:11:16 Unraid kernel: NVRM: GPU 0000:05:00.0: RmInitAdapter failed! (0x24:0x72:1436) Dec 21 14:11:16 Unraid kernel: NVRM: GPU 0000:05:00.0: rm_init_adapter failed, device minor number 0 I have read that sometimes that rminitadapter failed error could be related to power. I am using the 6pin Pcie power cable from my PSU. I have also tried a 15pin Sata power to 6pin pcie adapter and saw the same error. Attached are my diagnostics. Any suggestions? unraid-diagnostics-20231221-1417.zip

I followed @S1dney post that is a work around. It DRASTICALLY reduced my writes. I would recommend you check it out. I can now sleep easily as night. One thing that confused me was that after I rebooted I didn't have any docker containers listed. I then had to select Add Container and then choose the template from the drop down. It was quick and only took about 10 minutes to re-add about 20 containers that I had set up previously.

Thank you for your help, and you're right, I don't want to derail this thread from the issue at hand. If anyone is curious, I was able to get it working after @S1dneys insight. To get it to work, I had to take the original rc.docker file from 6.8.3 and modified only the start_docker() with the updates. That restored the proper networking after a reboot.

This is great! I implemented your work around and re-added the containers from the templates like you suggested and the writes are drastically reduced! One thing I did notice and wanted to see if I'm crazy. Making the change to not use the docker.img blocks network access from the host to the macvlan containers even though I have it set to enabled. For example, my Unraid is 10.0.0.2 and my Plex container 10.0.0.5 is on the custom macvlan adapter (br0). While not using the docker.img pinging from unraid to plex, it never replies and get destination host not reachable. If I remove your work around and docker uses docker.img. Unraid can ping the Plex container. It isn't the worst thing, but I do have some reverse proxies that aren't working until I get this figured out. Any thoughts?

Thank you for the explanation. That makes sense. I will give that a try!

I tried following your tutorial, but after reboot it tells me I don't have any docker containers installed. It shows me the docker service is running. Checking the logs I can see it created the softlinks. If I remove that entry to copy the rc.docker file in the /boot/config/go directory everything works again after a reboot. However, then it puts it back to original and uses loop2.

yep.. I was just doing some math from my post above and came to the conclusion it made no difference and I'm still screwed...

EDIT: Nevermind... still screwed. I posted earlier that I'm affected by this issue as well. I am using a btrfs unencrypted cache pool with 2x 1TB Samsung 860 EVO SSDs. One thought occurred to me that I wanted to run by everyone. For your containers that you have running do you have them set to /mnt/cache/appdata or /mnt/user/appdata? I had all of mine set to /mnt/cache/appdata. Every time I would install a new container I would change the variable path to use cache. I just went through all of my containers that I have running and switched them to /mnt/user/appdata. Right now I'm at 2gb writes for loop2 writes after 10 minutes. I need to do more testing and see if this is any better. I'm grasping at straws as my SSDs are trashed now. I currently have 25 containers running: 1activegeek/airconnect grafana/grafana homeassistant/home-assistant influxdb:latest jlesage/filebot jlesage/filezilla jlesage/firefox jlesage/handbrake linuxserver/ddclient linuxserver/duckdns linuxserver/letsencrypt linuxserver/mariadb linuxserver/nextcloud linuxserver/plex linuxserver/unifi-controller modenaf360/youtube-dl-nas oznu/homebridge portainer/portainer saspus/duplicacy-web spants/mqtt telegraf:alpine

I'm having this issue as well with tons of writes to my Cache Pool from loop. Unfortunately, the warranty on my SSDs are now shot with just over a year of run time. Power on hours: 10373 Total LBAs Written: 1972163391415 - 918 TB unraid-diagnostics-20200416-1127.zip