• Posts

  • Joined

  • Last visited

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

btrcp2000's Achievements


Apprentice (3/14)



  1. Usually when I read enough things will beocme clear, but that's not happening in this case after several weeks. I have never used a vpn before, but have recently been playing with mullvad. FIgured I should start with wireguard since that's where everything seems to be headed. I am able to use the mullvad app on windows and android so far, testing with, so I understand that part. I would like to create one tunnel through which certain devices, dockers, and VMs would travel when I need VPN, but other work/school computers need to stay out of it. I have UNRAID with plenty of resources, and a Ubiquiti Dream Machine Pro. In my mind, all I'm looking to do is set up a connection through Mullvad (or something better, i'm open to suggestions), and the point stuff at it somehow, but with easy access to change the exit point location if needed, as sometimes Mullvad IPs get blacklisted, which I assume happens to all providers from time to time. I have seen plenty of great tutorials by SPace INvader one, Capt Chaz, Juan M, etc, but I can't quickly discern which approach would be best for this. Or if I'm already way off course. Am I better off by building this tunnel inside a docker, UNRAID-hosted VM, or the UDMP itself?
  2. @Cpt. Chaz I am attempting to use an Nvidia ShieldTV (android TV device) per Cpt. Chaz's wireguard Mullvad VPN-inside-a-VM setup explained here ( The Mullvad app allows split tunneling, but requires that apps be excluded one by one, which is literally like 300 when you include all the various android components. Would be great if you could "exclude all" form the VPN then add back one by one, but they see it very differently. I have recently been informed that an app called "OpenVPN for Android" is easier to use, and I'm fine with these devices using OVPN rather than wireguard. Ideally, I can simply point these specific androidtv apps to the VM I have already built. Assuming this is possible, how would I modify the mullvad conf file within the OpenVPN app so it sends traffic through the mullvad VM? Or is that not how it works, I can see that being redundant too? Or would I need a second VM setup as OpenVPN? Ultimate goal would be to route multiple LAN tv devices through the VM so as not to run into Mullvad's 5 device limit, so I am hoping that the solution is to configure the the OpenVPN app on each androidtv device to only send specific apps through the VPN, then point it at the VM dns to tunnel from there, rather than having it go straight to the internet.
  3. If this isn't getting attention anymore is there a better way? I am not getting anything at the early step of setting the static ip using either /etc/network/interfaces or the updated one in the comments /etc/netplan/50-cloud-init.yaml. This happens on the latest version of Ubuntu Server as well as the 18.04 version. All I get is a blank screen where I can type whatever I want and save it I guess, but nothing pre-populated like the output in the video Obviously very new to ubuntu, so all I can really do is follow paint by number instructions. Where am I going wrong?
  4. Two instances of this in the past few weeks, once when upgrading UNRAID and the subsequent reboot, another when shutting down to add more RAM. Server will get to the point of full shutdown, then all drive LEDs light up and it just screams in one long steady beep. SuperMicro says this indicates a heat issue, but I don't see any alerts in IPMI, and the heatsinks don't feel any warmer than usual. It will fully shut down with press of the power button, then reboot. First time it fails with a complaint about mounting a boot device. Hard press again, it boots up as if nothing ever happened. While running, I don;t notice anything out of the ordinary. Right now, it's running like a top, in fact I am typing this on a Win10 VM. Am I about to lose something? Flash and important data is backed up via crashplan as recently as last night, so I should be okay worst case scenario, but is this something I need to address or just wear ear protection when I reboot?
  5. Set this up a few weeks ago and have noticed some strangeness that began immediately. None of this makes sense, but it didn't start happening until after Adguard. I set up the docker, gave it its own IP, and set my Ubiquiti USG to use that same IP as its DNS, with as a secondary DNS. I know it works as I can of course see it in the Adguard web page, and it does block the ads on specific sites I was hoping for. I have not done anything custom yet. Here's the weirdness: Reddit cannot connect in the mornings, around 6am, for some reason. It's fine other times Google maps is intermittent, sometimes it works, sometimes it can't connect either. Netflix comes and goes too on specific computers, but not Android devices (shieldTV) These are not issues outside my own network. IE, when reddit loses connection in the morning, I just shut off wifi on y phone, and it connects immediately via the cell network. Also, what is the difference between using the docker and just setting my routers DNS to Adguard's address? Customization? TIA
  6. Just adding that I am seeing this too. Server doesn't crash, but dockers become very sluggish, some are dead in the water. VMs somehow are not impacted, I guess due to CPU pinning? But these spikes seem to hit all CPUs one at a time, like OP says. EDIT: Upgraded to 6.8.2 last night, will keep an eye on it and report back
  7. Another spike tonight. Tried killing Crashplan, but it continued. What else is going on here?
  8. Ok, I will watch for that next time. In the meantime, what about this from the log? This would have been a recording done by SageTV that Plex then looks to index. Not sure what is going wrong here though.
  9. Can someone take a look at these diags and tell me what freaks out and spikes half to 3/4 of my CPUs and brings unraid to a halt periodically? It eventually recovers, but I can't pinpoint what is happening at the time. I suspect Plex, but not sure why. TIA
  10. I have an old beater monitor connected to my srver's onboard VGA port so I can monitor server reboots or see what is happening under the hood when things go haywire. I have noticed that it seems to "sleep" once my VMs start up, and I won;t see any output on the monitor until the next reboot. To be clear, this is to watch POST events, NOT passed through to a VM or anything. I did briefly accidentally set it to pass through to a now-deceased VM. Not sure if that has any impact. How can I get VGA output to stay put? Diags attached
  11. But what would have happened to the data on the other drive that never showed any issues? Dockers survived, so why just the domians folders?
  12. unraid-diagnostics-20191129-1500.zipSorry, I should know better by now. Attached. That rings a bell though. On first boot after re-running cables, it lost one of the cache drives. On another boot, the missing drive showed as unassigned, so I assigned it as cache thinking it would just sync up with the other cache drive, as that's the point of having two (right...?) Working fine now showing dual cache drives. Still need to learn best practices to offsite backup (this weekend's project)
  13. I'm not heartbroken because my VMs were a mess anyway, but now that the important ones are up and running better then ever I don't want this happening again. Why would one folder have gotten blown up? It's as if there was never anything in the domains folder, and there were 6 VMs the other day. What is the best way to backup VMs? I'm seeing lots of options, currently I just have Crashplan syncing that folder, but I'm guessing that's not a good idea because it doesn't know to shut them down?
  14. Also, you mentioned virtualizing with the onboard. I did that briefly but it caused numerous issues, although my onboard GPU is nothing to write home about anyway
  15. I am by no means an expert, just sharing what I've gone through so hopefully someone wiser than me will chime in to help. My final determination was that something in my system was not going to permit the passthrough of two GPUs period, regardless of manufacturer. I battled with nvidia for months (search for code 43) before trying the HD5450 which was plug n play, just check the box in the VM config, and its still working fine right now. A second GPU in a second windows VM would only come up as Microsoft Basic Display Driver at 800x600 resolution with no ability to adjust, Manufacturer installer programs (inside the windows VM) wouldn't recognize the hardware either, tried with 2 nvidia and 2 amd. Best advice from my limited experience is stick with AMD, as for some reason Nvidia actively disables VM passthrough if I read correctly. Some folks here have gotten it to work, but few of those threads have any definitive solutions, at least none that worked for me. If you can live with an external GPU, just plug it into onboard USB and check the box. Plugging it into an already-passed-through usb card worked but would only allow the external to be a secondary extnded monitor, with VNC having to be primary, which made my desktop monitor useless. There was no option to mirror the two.