tbrasser

Dear unRAID community, I've been running unRAID for a few years now, and most of the time, everything works as expected. However, I keep having weird DNS issues when I'm trying to route all my internet traffic through a VPN using my (opnsense) router. My reverse proxy setup works fine (always), but seemingly at random I fail to resolve domainnames from containers / the unraid host / sometimes other devices on the network. I've been reading up on the Docker custom networks and how it works internally, however, I've been unable to get it to work. Current situation: public-ip-->[ISP Modem/Router]192.168.178.1-------192.168.178.10[OPNSense Router]10.0.0.1/24----------Rest of my network (unRAID server, WIFI-AP's, Switches etc) The OPNSense router is in the DMZ of my ISP-provided box (that I unfortunately can't put into bridge mode) I have allowed private IP ranges on WAN on my OPNSense router I have set 8.8.8.8 and 8.8.4.4 as the only DNS in the OPNSense router, this is what all clients get via the DHCP server that is running on the LAN interface. I have disabled any type of DNS resolver/forwarder for now. This DHCP server gives out adresses from 10.0.0.64 till 10.0.0.127 I have 2 interfaces in my unRAID box, eth0 and eth1, I have configured unRAID as follows: eth0 has the ip 10.0.0.10 via DHCP (static lease), bridging is disabled eth1 has no ip, bridging is enabled (br1) Docker custom network on br1 is 10.0.0.0/24, default gateway 10.0.0.1, no DHCP. I spin up every docker container with a specific IP on br1 (between 10.0.0.128 and 10.0.0.254) and I provide the --dns=8.8.8.8 flag. The problem I keep facing, and the reason I went and configured everything as above, is that whenever I make changes on how devices on my network access the internet (over VPN or not etc) I get weird DNS issues, either my unRAID host can't ping google.com, or my docker containers are unable to resolve domainnames. What I'd want ideally: I can't begin this list without stating that I'm trying to get rid of this DMZ / double-NAT situation, it sucks. I want a simple network (for a start, can do vlans later for iot/guest things) All devices on my network should be able to resolve internal and external domains. I like my current setup of using my /24 subnet as follows: .1 for the default gateway, .2-.63 for static leases, .64-.127 for dynamic leases, .128 and above for docker containers running on the unRAID host. Here are some screenshots and outputs of various commands: - unraid gui settings related to networking and docker - opnsense gui settings related to dns/dhcp and overview - ping and dns results from my laptop (10.0.0.20) - ping and dns results from the unraid host (10.0.0.10) - ping and dns results from a container started with --dns=8.8.8.8 on br1 with 10.0.0.181 ip Can someone help me debug this? I haven't been able to get a stable network for ages now, and I have no clue what I'm doing wrong. server-diagnostics-20181027-1759.zip

1. Pinging my router works fine: root@Server:~# ping 10.0.10.1 PING 10.0.10.1 (10.0.10.1) 56(84) bytes of data. 64 bytes from 10.0.10.1: icmp_seq=1 ttl=64 time=0.255 ms 64 bytes from 10.0.10.1: icmp_seq=2 ttl=64 time=0.294 ms 64 bytes from 10.0.10.1: icmp_seq=3 ttl=64 time=0.143 ms 64 bytes from 10.0.10.1: icmp_seq=4 ttl=64 time=0.202 ms ^C --- 10.0.10.1 ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3095ms rtt min/avg/max/mdev = 0.143/0.223/0.294/0.058 ms 2. Oh yeah, it's still on a /24 subnet, the /25 is the DHCP pool for the br0 docker network: Subnet: 10.0.10.0/24 Gateway: 10.0.10.1 DHCP pool: 10.0.10.128/25 (128 hosts) Can't ping google's DNS servers either. However while typing this comment I took a look at my pfsense config again, and found that the only thing different between the unraid server and the laptop I'm typing this comment on now is that I've had my laptop configured to go through an openVPN interface, I just tried to add the unraid server to that alias list and now it's connecting again. weird. So it's safe to say that the problem is not at the server, it's in my router config. Thanks for the reactions anyway, I'll be reconfiguring my router/firewall now to tackle the problem.

My unRAID server can't access the internet. I've searched the forums & checked my network but I can't find anything wrong. Pinging google / github / etc from console doesn't work. I run a physical pfSense router, which runs a DNS (I use dnsmasq blocking, but have that disabled to rule out that that was the problem, backend DNS is 1.1.1.1 ) Pfsense box: 10.0.10.1 LAN: 10.0.10.0/24 DHCP pool: 10.0.10.64-127 unRAID server: 10.0.10.10 docker network: 10.0.10.128/25 This started happening after I updated my pfsense firmware, but no other devices on my network seem to have any issues. Diagnostics are attached if anyone wants to take a look, hopefully it's something simple I've overlooked. server-diagnostics-20180805-1511.zip

you can dump the bios from the unraid command line as per the video's instructions. Works for me, I passthrough my gtx 970 to windows VM and its the only graphics adapter that is installed.

Also, with looking glass, you'd still need to pass through an entire gpu to a VM (for example windows), what you'd then could do (not with unRAID as jonathanm correctly pointed out) with a linux host is have the framebuffer of the windows gpu copied to the frame buffer of your linux host so you get a kind of picture-in-picture, and you only need 1 cable.

I narrowed it down to my /var filling up with syslogs. Turned out my VM setup with primary nvidia passtrhough is giving me BARs warnings continuesly. Not starting any VM's on boot made everything work again. So it didn't have anything to do with this plugin in the end, got LibreElec 6.5.0 running now! Thanks for the help anyway, and if someone knows how to stop uninteresting log entries from filling up the partition (or is familiar with the type of error I got, posted in the 6.5.0 release thread) feel free to post it (A)

Thanks, I just found it as well. Not sure what is causing it, any ideas on where to look for? So after some googling I didn't get much wiser other than that those log entries don't indicate anything serious. However they do fill up my /var .... It seems to be a result of my NVIDIA GTX 970 + Some linux kernel changes that have happened in the past. To pass through my gpu (it's the only one I have) I use an extracted bios that since 6.5 I can point to without modifyin the xml. I have 1 VM on autostart which probably resulted in above spamming in the logs. I can confirm that if I don't start any VM I don't get these log entries, and everything is functioning like it normally would. However I would like to start VM's with my GPU without getting this behaviour. Is there any way I can reduce the logging or any other fix?

I'm not sure whether I'm posting in the correct thread. But I've started noticing these issues when trying to upgrade from 6.5.0-rc to 6.5.0 (DVB edition, but now I'm on 6.5.0 stock). I can't seem to install (some) plugins, I get tar errors. I think I've narrowed it down to the fact that my /var is full? Can someone take a look at my diagnostics to see if that's whats the problem. (and if it is, how can I correct it)? I've also posted in the unraidDVB thread, but I don't think my problem is specific to that plugin. (I've also had CA failed to install after I noticed my APPS tab was missing, however it showed back up on 6.5.0 stock) server-diagnostics-20180315-1617.zip

I was on 6.5.0-rc6 and when I wanted to upgrade to 6.5.0 weird things started happening. Initially I got an error on the DVB plugin page. So I tried to uninstall the plugin which said it wasn't installed in the first place. Then I tried installing it (again), but I got errors when it tried to extract the tar file (errorcode 2). What I did is upgrade to 6.5.0 stock using the normal upgrade path. And after that I tried installing the plugin again. Same error: plugin: installing: https://raw.githubusercontent.com/linuxserver/Unraid-DVB-Plugin/master/plugins/UnraidDVB.plg plugin: downloading https://raw.githubusercontent.com/linuxserver/Unraid-DVB-Plugin/master/plugins/UnraidDVB.plg plugin: downloading: https://raw.githubusercontent.com/linuxserver/Unraid-DVB-Plugin/master/plugins/UnraidDVB.plg ... done plugin: downloading: https://raw.github.com/linuxserver/Unraid-DVB-Plugin/master/archive/UnraidDVB-2018.03.14.txz ... done +============================================================================== | Installing new package /boot/config/plugins/UnraidDVB/UnraidDVB-2018.03.14.txz +============================================================================== Verifying package UnraidDVB-2018.03.14.txz. Unable to install /boot/config/plugins/UnraidDVB/UnraidDVB-2018.03.14.txz: tar archive is corrupt (tar returned error code 2) ---------------------------------------------------- UnraidDVB has been installed. CHBMB/Squid Version: 2018.03.14 ---------------------------------------------------- plugin: installed Is there anything I can do to fix this. Before the upgrade I had the same issues with CA, the apps tab was not showing anymore (6.5.0-rc6) but it now is back on 6.5.0.. Here's a screenshot of me uninstalling the plugin after it failed to correctly install:

I'm eager to try it. Got a used dell T7500 and got some 8TB HDD's and 2 SSD's show up. But I'm wondering at wat speed they're connected.

12GB of DDR2, why? Because I had 4 and that was not enough to run a VM + some containers. And it's cheap. @electron286 You should check-out PKC: https://github.com/croneter/PlexKodiConnect

My "best" gues of what could've gone wrong: * corrupt download the first time around (didn't check the hashes or anything) * my unraid webinterface has a history of being slow, with unraid DVB edition plugin timing-out for me even before 6.4, but usually after some weird sequence of refreshing and stopping the loading of the page I got it working. (Which means it got stuck after downloading the list of available images), refreshing plugins to update or docker containers always takes ages (with around 15 plugins to check and 25 containers). Maybe with the new webserver changes (I read something about nginx) this "workaround" I used to have to go through also failed. (My hardware is pretty old, dual-core i3 with hyperthreading, 12gb ddr2 ram, no dedicated gpu, and probably a realtek gigabit LAN interface that was never ever intented for any kind of server-y workloads.) That being said, as long as plex isn't transcoding there are few hickups with the rest of the system.

I tried that now. Also directly on the server (booted in gui mode). Still no luck. Im giving up for today. I'll post back here when I have found out more. EDIT: I once again copied all the images to /boot manually and everything works! (well except for the DVB plugin, but the manual update worked, I can use my tuners now!) Thanks for both of your time!

I tried firefox on the IP, but again, same thing. You mean the UnraidDVB-2017.06.16.txz file to where?

Unfortunately I get the same behaviour as before I went through unRAID versions (see screenshot on the previous page): Edit: Occasionally I seem to be able to also produce the following warning, not sure whether it's related: [Violation] Added non-passive event listener to a scroll-blocking 'mousewheel' event. Consider marking event handler as 'passive' to make the page more responsive. jquery-2.1.1.min.js:3

I'm feeling like I'm missing something really simple here... Any incompatible / buggy os/browsers? I'm on: Version 59.0.3071.109 (Developer Build) Built on Ubuntu , running on elementary 0.4.1 (64-bit)

That gave this within about 15 seconds, (via ssh) so seems to function normally: <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN"> <html> <head> <title>Index of /unraid-dvb-rc</title> <script src="/cdn-cgi/apps/head/SnTkGghPCkY6gCE0hGVDVpppPno.js"></script></head> <body> <h1>Index of /unraid-dvb-rc</h1> <table> <tr><th valign="top">&nbsp;</th><th><a href="?C=N;O=D">Name</a></th><th><a href="?C=M;O=A">Last modified</a></th><th><a href="?C=S;O=A">Size</a></th><th><a href="?C=D;O=A">Description</a></th></tr> <tr><th colspan="5"><hr></th></tr> <tr><td valign="top">&nbsp;</td><td><a href="/">Parent Directory</a> </td><td>&nbsp;</td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-3-5/">6-3-5/</a> </td><td align="right">2017-06-06 16:33 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-4-0-rc1/">6-4-0-rc1/</a> </td><td align="right">2017-06-13 10:08 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-4-0-rc2/">6-4-0-rc2/</a> </td><td align="right">2017-06-13 10:11 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-4-0-rc3/">6-4-0-rc3/</a> </td><td align="right">2017-06-16 11:35 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-4-0-rc4/">6-4-0-rc4/</a> </td><td align="right">2017-06-21 22:55 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-4-0-rc5/">6-4-0-rc5/</a> </td><td align="right">2017-06-22 21:20 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><td valign="top">&nbsp;</td><td><a href="6-4-0-rc6/">6-4-0-rc6/</a> </td><td align="right">2017-06-27 21:04 </td><td align="right"> - </td><td>&nbsp;</td></tr> <tr><th colspan="5"><hr></th></tr> </table> </body></html>

So I went back to 6.4RC6 via 6.3.5, plugin installed fine, but same old, same old.

No, I have a static IP from my ISP and I'm in control of everything except my modem. (Running ddwrt on my gateway/router/wifap, with ip 10.0.0.1), then some static IP's incl my server till 63, then 64 till 127 are for DHCP. On unRAID I have configured the network br0 (which since 6.4 is of type macvlan) for docker, they all have IP's from 10.0.0.128 until (ultimately) 254). I also edited the route masks from 10.0.0.0/24 to 10.0.0.128/25. (Which is working perfectly by the way, I hoped I would also get host connectivity to my docker containers this way since everything below 128 would get routed to my router and everything above it to br0, but no luck)

I'm going to disable docker (because I'm using the new macvlan br0 stuff), then copy all the images from the "previous" folder back. (Which should get me to 6.3.5), then disabling as many plugins as possible + remove leftover files and folders if there are any, then going official to 6.4RC6 and do your suggestions. I'll let you know when I get any further. Thanks so far!

I upgraded from 6.3.5 DVB Edition to 6.4RC6 via the official way (with the new plugin). After that I updated all my plugins. (Then the Unraid DVB plugin wouldn't load the images so I tried to copy them in manually. ) How can I check whether the images actually start downloading (and maybe that times-out) or whether it hangs before that? Here are the contents of my /boot (and /boot/config/plugins) folders, maybe it shows something wrong:

Hmm. Then I have no clue what went wrong with the manual update. I'm still trying to install any build via the plugin, this is what I get in the logs: Jul 20 15:42:28 Server nginx: 2017/07/20 15:42:28 [error] 5227#5227: *100072 upstream timed out (110: Connection timed out) while reading response header from upstream, client: [removedclientip], server: , request: "POST /plugins/UnraidDVB/include/exec.php HTTP/1.1", upstream: "fastcgi://unix:/var/run/php5-fpm.sock", host: "server", referrer: "http://server/Settings/UnraidDVB" and this is what I get in chrome:

I tried that, I already had the latest version, but I removed it and installed it again. It still hangs/times-out at pulling the builds. So for the files I pulled manuallly yesterday, I got them from http://files.linuxserver.io/unraid-dvb/6-4-0-rc6/libreelec/ Is there any difference between those and those in http://files.linuxserver.io/?dir=unraid-dvb-rc/6-4-0-rc6/libreelec ?

I upgraded to 6.4RC6 from 6.3.5 and tried to follow your instructions. The settings->unRAIDDVB page keeps fetching images however and never stops loading. Then I copied all the non-checksum files from http://files.linuxserver.io/unraid-dvb/6-4-0-rc6/libreelec/ to my flash drive, and now I can see (while the page is still loading) that the currently installed build is LibreElec version 1.2.1, but there's no DVB Kernel detected. (So also no devices, and therefore nothing showing up in my TVHeadend container (with device passed through).) I've tried to trace back why the page doesn't finish loading, but all I can see are (nginx) timeouts of exec.php in the plugin directory...

According to this page it should be possible? https://ubuntuforums.org/showthread.php?t=1400905&page=2 Sorry for the massive bump, but I was searching for this exact thing. Otherwise I'll need to set up a VM just for mopidy...