TheOgre

Any idea? This is happening every single reboot. Just happened today, only workaround that works is Docker service > Disable, host access > disable, host access> enable, docker > enable. This is severely frustrating to have to do every single reboot. Please help lol

Hello all, been suffering with a really weird issue and can't seem to uncover what is going on. Each time I reboot my server, only SOME of my docker containers will start, it seems those using br0 all fail with Mar 28 12:44:13 Server rc.docker: nextcloud: Error response from daemon: failed to create the ipvlan port: device or resource busy Mar 28 12:44:13 Server rc.docker: Error: failed to start containers: nextcloud Mar 28 12:45:04 Server kernel: br0: Device is already in use. The only fix that works for me is, to disable the docker service, disable "Host access to custom networks", Apply, then enable "Host access to custom networks" and start the docker service, and then the containers start fine. I am using IPVLAN. Not sure where to go from here as there are no other log entries for it and googling produces no results. Any help? It's very frustrating having to do this on EVERY reboot of the server, and it seems like such a simple problem. ogreserv-diagnostics-20240328-2036.zip

Just would like to add to this thread that this has been an issue for a few people for a while including me and it is still occurring. No orphaned containers and nothing out of the ordinary here.

Just wanting to tag onto this thread, even after switching to ipvlan, the plugin is still detecting macvlan traces. Any way to diagnose why or anything else I need to do?

Just wanted to post an update here, I was also having this issue on 6.11 and just updated to 6.12.1 and was hoping this was one of the fixes, I had about 10 containers to update and using the update all function, I had the loop occur, only 2 loops and then it stopped but it still seems to be an ongoing issue. There seems to be numerous threads about this but haven't seen any unraid staff in any of them, anyway to make this more known as an issue? @Squid

Just updated to 6.12.1 and still having this issue. Anyone know what's going on or any fix for it?

So one issue I am running into is exclusions, I got the exclusions for each docker container figured out, but one big issue for me is the exclusion of the /.Recycle Bin/ folder from the Recycle Bin plugin. I find this plugin necessary for accidental deletion of files/folders and had it excluded in the old CA backup. I see under each docker container configuration there is an exclusion list and I can select basically any folder on the system from there, but will that actually propagate to the main backup? Seems a bit confusing as is. Any advice would be appreciated, thanks!

So I am actually running into a very similar problem. It seems that having more than just a few dockers running will crash my ISP modem. What's weird is that this hasn't been an issue for almost 3 months, where it last popped up then mysteriously disappeared. I have a few dockers running and everything is stable, but starting anymore than this will cause the modem to crash. Everything else in the network stays functional, all the Ubiquiti gear, routers, switches, but it's the ISP modem that is crashing. Following this thread to see if there's any helpful advice given here.

Was searching for the solution to this as well. The threshold is actually under Settings > Disk Settings > Default warning disk utilization threshold (%) Unfortunately this applied to every disk connected to the server, so your array and unassigned drives will be affected.

The docker image changes only if dockers add or update new images to it, this does not happen during normal operations. A memory leak will only cause RAM to increase, not the docker image.

This isn't related to memory usage but the size of your docker image. What's most likely happening is a docker is updating, at which point it expands the size of the docker image, then once completely updated, it clears the unnecessary parts of the image. Can resolve this by increasing the size of the docker image.

Very strange they vanished.... However yes, it should be the same process to get them all back if you haven't already.

I'm not sure if this will work but go to http://YOURSERVER/Docker/AddContainer as a URL and see if it will let you recreate the container from there, after that the others should stop rebuilding.

When you say "auto-rebuild" are you saying the page keeps refreshing and the container IDs keep changing on some of the other containers? That usually happens AFAIK when containers get names mixed up with their networking or something similar. If there are multiple containers doing it, might be best to restart the docker service and see if it continues or resolves.

On the docker page at the bottom, create a new container, then select the template drop-down and select binhex-qbittorrentvpn and it should repopulate all of your variables.

Definitely have to make sure none of the ports overlap as you'll only be accessing them via the HOST IP, just make sure the first 2 variable ports are the same for each container (UDP/TCP respectively)

Odd that it's worked in custom bridge mode for over a year and a half and suddenly has had issues, changing this setting has fixed everything.... Thanks!

Just as an update to this, the issues seem to be randomly getting worse and only with this container. Privoxy is now not allowing connections even though the log shows it is listening on 8118, radarr, sonarr, nor jackett can connect to privoxy, and now they're having issues connecting to qbittorrent just like I am on my local LAN, using WireGaurd allows me to access the webUI to qbittorrent. Have tried removing everything from LAN_NETWORK except 10.0.0.0/24 and it does change the fact I cannot access anything via WireGaurd, but it's still not allowing LAN access. Have also deleted and reconfigured the entire appdata folder. As a side note, is there a way to update the docker template in the unRAID UI? It seems the only way I can get an updated template is deleting the container, then redownloading it from the Apps tab.

unRAID is on 10.0.0.35, qbittorrent is 10.0.0.36 on br0, accessing it from a PC or mobile produces same result, so we'll go with mobile for now. Phone is 10.0.0.15 static and cannot access, 10.1.0.2 on WireGaurd and has access. I would like to confirm that I have no changed any LAN_NETWORK settings or configuration to qbittorrent other than updating the docker container, same for WireGaurd, it's been a set it and forget it installation as far as configuring since I had it working. An update occured at some point that broke it for me a while ago, just got around to diagnosing it.

The value now is simply 10.0.0.0/24,10.1.0.0/24 The first is the actual LAN the server is on, and the second is the WireGaurd pool. Only when connected via WireGaurd can I access the webUI. I would also like to add I tried 10.0.0.0/8 but that resulted in some pretty container breaking issues as it was trying to pull API requests from a local IP at that point.

For some reason, this nor longer works for me, in a very odd way I might add. Originally, all 3 subnets could access the WEBUI, 10.0.0.0/24,10.1.0.0/24,10.2.0.0/24, however now, I can only access the WEBUI via the 10.1.0.0 subnet, which isn't even the main lan subnet. Any particular reason this may be happening? Super annoying not being able to access it without having to connect via VPN

Okay so, after following your suggestion @Squid, it made me think to try a Live ISO and see if the network issues persists with a known working live drive. Well, it did not, and this led me down the road of trying the network stuff again. Turns out, I have a malfunctioning network switch. It seems to be passing SOME requests, but dropping everything else. After running a very long cable to the main ap, everything works just fine in unraid. Thanks for the nudge that got me to the solution!!

I have not made any tweaks to networking via that plugin, however, I just tried as you said and disabled it but no changes. I also tried booting in safe mode with no plugins and it is still not working. The very odd thing is, monitor my dhcp server, it does get the request and is assigned an IP, br0 then holds onto that IP, but eth0 drops it and gets the "device interrupt 20 memory [ADDR]" note underneath.

Having very odd issues with networking on this machine. Upgraded my main computer and repurposed my old one to replace my existing unraid server. Everything went flawlessly for about 2 days, now, networking is not working with no changes to the server or my network stack. ifconfig from the terminal shows eth0 with 10.0.0.35 as well as in the webUI (accessed via booting GUI mode), but diagnostics shows that it does not have an IP. An interesting discrepancy. I am uploading 2 diagnostic files. The first is when I initially accessed the server after networking went down, which shows "e1000e 0000:00:19.0 eth0: Detected Hardware Unit Hang" as the only issue I can see. However, these issues are showing up even when the server was still working fine and was accessible. Have tried swapping cables, different ports on the switch, rebooting the entire network stack. Changing DHCP settings. Nothing seems to change. I have noticed that the first minute or 2 after booting in GUI mode, I can access the network/internet from the server, then it suddenly goes down. The ui still shows it has an ip address, and if I unplug the network cable, it shows the link goes down, plug it back in, it comes back up full 1000mbps full duplex, however, still doesn't have access and ifconfig at that point shows it has no IP either. I am completely unsure what is happening, especially considering this was working just fine for 2 days straight. Is it possible the on-board NIC is dying? I can order a NIC card and put in, just will take some time to get here. Any help would be greatly appreciated. System specs: 4790k ASUS Z97-a Onboard Intel i218-v NIC 24gb DDR3 1866mhz ogreserv-diagnostics-20220129-1412.zip ogreserv-diagnostics-20220129-1547.zip

I have tried this docker container twice over the past few months and neither times has BlueIris actually opened. The docker container starts fine, i am able to see the noVNC feed and access the desktop, but nothing ever comes up, when right clicking, and selecting anything in the BlueIris applications tab, none of those commands do anything, and nothing shows in top either. Does this container actually work without pre-configuration? Or am I completely missing something that allows BlueIris to start? Any help would be appreciated as I am trying to dump Zoneminder for something that works and doesn't seem to have constant issues.