Defylimits

Well I had the same issue, thanks Celsian, Have to agree with Meep tho, very counter intuitive, been bugging me since uprgraded to 6.12.2. Weirdly it didn't get resolved when I tried to downgrade!

This Solution from Celsian worked - Thread - https://forums.unraid.net/topic/140858-vms-on-br0-cannot-access-the-internet-after-6120/#comment-1283343

Ok bit of an update, I reverted to 6.12.1 but appears to have made no difference However I did notice my Windows VM can access docker webpages that are either on my custom docker network "dockernet" and dockers that are assigned to the "host", but not any of the dockers on br0, so there appears to be something blocking the VMs from communicating with the main network. So I'm going to guess its something to do with the routing table, but not sure how to fix that tbh.

So I have 2 VMs I run constantly, my HAOS and a Windows 10 VM. They have both recently lost connection to my network, and it seems to be after I did an update 6.12.2. I do have VLANs running, a Server VLan (100) 10.10.10.0/24 and an IoT Vlan (20) 192.168.20.0/24. With both being on the 100 Vlan, and the HAOS having a second connection to the IoT VLan. The Unraid interface is on the 100 Vlan as well and is the default Vlan for the port the server is plugged into. I have tried updating the Network Source, from br0.100 to br0 and even virbr0 but no connection. Have also tried chaning the Network model to virtio-net from Virtio io but not luck then either. Weirdly I can access either of the IP's and the Window VM shows no connection, however in my Ubiquiti router I see all 3 connections and fixed IP's from the 2 VMs even after refreshing the MAC addresses from the Network connections as well. Am really confused, all the network connections for Dockers are working, some of which are host, others bridged and others on custom network. Not sure what the problem is or what to try, anyone have any idea? Diagnostics attached btw. tower-diagnostics-20230706-1638.zip

Idiot, didn't Stub the graphics! Tools >> System devices - Select IOMMU Group with Graphics card in and scroll to the bottom clicking "Bind Selected VFIO At Boot"

Looking at the Syslog I have these errors around the time having the issue? Am I missing something?

So I've got an interesting issue, if I assign a graphics card and then start the VM, my VM Manager seems to crash. When I click on the VMs tab or on VM Manager under settings both pages from the Unraid GUI it will not load, I also loose access to the Dashboard page which will not load correctly. VM's already running still appear to be accessible when this is happening though and Dockers appear to be online? To get the system back working I need to reboot, but I can't do this from the graphics and have to go power off the server physically and then restart it that way. I've got three graphics cards in my machine, a GT710 for Unraid, a Quadro P600 for Plex transcoding and the 1070 I'm looking at passing through to my VM. I tried this on the Windows VM I had, just by changing from VNC graphics to the 1070 and experienced this crash a number of times, reseting back to VNC graphics and everything works fine. Looking something up I found it might be an issue running the OVMF BIOS, so I setup another Ubuntu machine with SeaBios, but just had the same issue. Wondering if anyone has any ideas? Btw hello again everyone, last time I was in the forum was back in March 2020, pandemic kept me away from home for some time so I didn't have a chance to fiddle and bugger anything else up until now!

Thanks! I hadn't realise that you could reorder the interfaces. So just an update for you, eth0 - 10GBe Network Card ---->> br0 (Dockers and VMs) eth1 - 1 GBe Onboard (No IP Address) eth2 - 1GBe Onboard (Port Down) So that's how I've got it working so far and changed my strategy. Now I want to have a second bridge conenction linked to eth1, so that I can assign some dockers just to use that port. However whenever I follow the setup in your guide below my docker image becomes orphaned. As soon as the docker is started with this - "Modify any Docker via the WebUI in Advanced mode Set Network to None Remove any port mappings Fill in the Extra Parameters with: --network docker1 Apply and start the docker The docker is assigned an IP from the pool 10.0.3.128 - 10.0.3.254; typically the first docker gets the first IP address" Not sure what the cause of this would be?

That's great thanks, how do I also access the network shares through this connection as well? I mean to be honest I'd prefer just to run everything off the 10Gbe connection if possible? Maybe with the 1gig connection for management access?

Hey eveyone, just wondering if someone can help me out as bit green with networking in general and on unraid. Just added a new network card into my server (10gbe as it was cheap) and wondering how to configure my dockers and VM's to use it. Currently I have eth0 - 10.10.10.208 - Motherboard port 1 - Unraid host connection, only connection I've been using up until now eth1 - Port Down - Motherboard port 2 eth2- 10.10.10.210 - Asus 10gbe network card - Connection I want to use for all dockers and VMs. I disabled bridging of eth0, and enabled bridging of eth2, and custom br2 was created. Reassigning the VMs to this bridge, with their IP addresses set to 10.10.10.211 and 10.10.10.213 appears to have got these to work. Now on to the dockers, that are runing on the Host network and so use the 10.10.10.208 address, however I want these to use the 10.10.10.210 address, how do I point these in this direction and thus use the 10gbe port? If I bridge the docker then I get something like this "172.17.0.3:9987/UDP --> 10.10.10.208:9987" (teamspeak), and if I use br2 like the VMs I get "10.10.10.209:80/TCP --> 10.10.10.209:80" (webservice). I feel like I need to create a seperate Host2 or Bridge2 network type that I can use to select when trying to assign a Docker, but am unsure how to go about this, anyone got any pointers?

I have a similar problem, the P5000 and P4000 (one in my server) are supported by this driver version (see snip)? My 710GT and 1080 both showed up in the Info page, and the 1080 disappeared when it was being used as pass through for my VM, but the Quadro P4000 does not show up on the Info page even tho it is showing up in the system devices page. Is there a way to get the UUID in a another manor and still try passing it through to the docker? Or will it not work at all if its not detected on the Unraid Nvidia page?

Right so a bit of an update, finally moved into a new house only after 9 months of waiting *sigh* Anyways I found that changing the Video cards to EFI mode from legacy allowed the machine to boot and all the cards show up in Unraid and have even pass through the 1080 I was planning for the machine. Score! However I don't seem to get Unraid on any of the display outputs that are linked with the machine. Does anyone have an idea why that might be?

Basically yes, I've tried most posts talk about putting two of the same cards in sadly. What would be cool is if Unraid can reenable the Pci slot when booted, but I'm guessing that is going to be a long shot.

Good point, just upgraded from v3.5 to v3.95 Still the same issue, and had a heart stopper at one point thinking I'd bricked it. That moment when you start it up again and nothing displays for ages!!! Thanks tho

So as my signature says I have a HP z820 workstation, and was looking to install additional graphics cards in it. Got a 1080 I want to put in and assign to a VM and a quadro 4000 I was to use for transcoding if I can get away with the two in my power budget. Got a GT 710 that am using for the unraid graphics card btw. Problem is that the HP workstation throws an error when there are more than one card is installed and then wont boot. I can get it to boot by disabling the PCI slot that the other cards are in. Would there be a way in Unraid to then re-enable these once the system has booted? Thanks in advance for any help / suggestions

Hey thanks, I never got to try this because I had installed another VM, but if I move them disk again (more than likely) will try this.

So further update on playing around with this, I did end up installing another windows 10 VM, works fine, until you unassign the disk in the configuration and then reassign it. DIdn't even move the Vdisk location. Anyone got any idea's?

So I tried assigning it to a newly created VM and still get the same issue, but thanks for the tip!

Also just as an FYI i'm on Unraid version 6.7.2

New drive is also through unassigned devices. This is my VM XMLserenityxml.txt TBH I also seem to have an issue with the disks showing up as 246G when it was assigned as 400G previously. Trying to change it I get the sucessful message displayed at the bottom but the value always reverts to 246G.

Hi there guys, Wondering if someone can help me, got a Windows 10 VM that is refusing to boot. I have recently moved the VM disk from one unassigned device (old 1TB spinner) to a new one (500GB SSD), using the Dolphin app to copy across the file. Then I've changed the location of the machines disk to point to the new vdisk location. Tried to boot up and get the attached boot screen image which has the statement "Press ESC in x seconds to skip startup.nsh or any other key to continue". Can get into the bios menu but there appears to be no disk located in there and still doesn't boot when I exit the BIOS. Tried to re-point the VM back to the original disk location and get the same error. Wondering what I can do to get it to work, any idea's?

56GB of ECC RAM, as one of my 8GB sticks died. I did have up to 128gb when I was running an VMware system in a HP DL360, but that as too noisey for home usage when I had to move it out of the Data centre it was in. Also when I migrated to UnRaid and never looked back!

Thanks for the quick responses! Already done! I'm assuming if a disk does now go *pop* then I'll lose the data on that disk, but will be able to recover the data from the other disks.

Morning All, So I've got a bit of a problem and wondering if anyone can help (although I suspect I already know the answer!). Have an array of 6 disks and recently had one have an issue and has been disabled (contents emulated), see the attached Smart information. Not sure how to read that if anyone could give me some pointers of what is wrong with it. Also attached the disk logs which make it look like its a sata link issue?? So the disk didn't have much on it and I've made sure that this has now been 'moved' off the disk onto others in the array, although there's no files on the disk in the explorer, the summary page says 4.02GB is used? Now I need to remove the disk to make sure that the rest of the arrays contents is protected. Here's the kicker, I'm not in the same country as the server, it currently lives at my families house (whilst I wait for my house purchase to go through but thats a different problem!), and my family have gone on extended holiday. I access the server by remoting into a VM run on the server, note this VM runs on a seperate disk through unassigned devices. As you can see I can't take down the array as I wont be able to access it to remove the disk and restart it, is there any way to remove the disk from the array without stopping it, or just blanking it and putting it back into the array to protect the parity? Or anyway to keep the VM up whilst the array is stopped. Running unraid version 6.6.7, I think I know the answer but thought I'd ask as its going to be a couple of weeks before I can get to/get someone to the server physically. Cheers Diskloginformation.txt WDC_WD40EFRX-68N32N0_WD-WCC7K0NNFX24-20190430-0720.txt

Thanks for the reply, Cheers, it would be good if you could just enter the network_ID's separated by a comma in the edit docker options Yes pretty much, I believe it is possible - https://support.zerotier.com/knowledgebase.php?entry=show&search-for=&article=ZWFhNWMyMTZjODY1ODcwNmFhZmJjYmRhN2I5MjRhOGQ_ Step 2 and 3 are easy to do. Its step 1 that requires some messing around with the IP tables within unraid/docker which would allow for the server to act as a NAT and forward traffic. (sorry not really a networking person!) Then you should be able to select on the client side whether you just want to connect to the zerotier network or use that network as a default gateway for all traffic. Unfortunately I've tried tower.local and it didn't work. Although I've had issues before with this on the local network it is on, so will have to look into the problem a little more. Thanks for pointing me towards the avahi daemon.