RAINMAN

It is a private VPN. I have 4 different VPS servers that I use for VPN. All have the same issue if I use a different one. Then again, they are all setup more or less the same but no other clients connecting to them have issues.

Thats correct.

unraid is the peer as I mentioned before. The remote VPN server has: ens3: 10.0.0.5 lo: 127.0.0.1 tun0: 10.8.0.1 wg0: 10.9.0.1 Local unraid uses: br0: 192.168.254.3 docker0: 172.17.0.1 eth0: some ipv6 address? lo: 127.0.0.1 bunch of vethxxxx: ipv6 wg0 10.9.0.6 No overlaps except the wg interfaces on both which is proper. Note, I am not using wireguard to connect from outside in. I am using it to route unraid traffic out over vpn.

Those IPs are not connecting to wireguard. They are only connecting to the nginx docker via my external IP. The one is a 192.168.0.x network the other i dont know since its not mine but i doubt their internal network address would be relevant when they connect via my external public IP. The only reference to wireguard in this is that when its enabled they cant hit my dockers, when its disabled they can hit them fine. Its like a routing issue where the reply to the request is going out over the VPN instead of directly back, if that makes sense.

The wireguard subnet is 10.9.0.x Local Lan is 192.168.254.x Remote IP example that doesn't work is 69.17.172.210 Remote IP example that does work is 140.238.153.159 Docker subnet 172.17.0.1?

I'm still getting weirdness connecting to my dockers. From certain external IPs they can't connect at all to any of my docker services unless i disable the vpn connection. Other external Ips are working fine so i am not sure how this would get messed up?

If I put by MQTT docker on a custom: br0 network and assign it an IP and port forward to that IP, then I can get my outside connection to work but all my internal connections are dead because I have to go through and re-program them all. (many are arduinos so its a PITA if I need to change the IP) Edit: it gets weirder, from a different webserver it does seem to connect fine. I'm not sure where it would be blocking traffic from that server only, and only when the VPN is enabled. Any suggestions where to look or what to look at?

Ok, maybe I had some field wrong, when I go through it again from scratch it works. There is no way to set DNS though. Perhaps there can be a field for additional custom options. My VPN will try and push 10.9.0.1 as its DNS for my external devices but when I am internal to my network I want to force DNS = 192.168.254.50, 192.168.254.30. I have VPN tunneled access, yes. I am not trying to send to the VPN though. Remote server -- 1883 --> Home IP (Router) Forward to -- 1883 --> 192.168.254.3 --> MQTT Docker. I wouldn't expect the VPN to have any impact on this route? If I disable the VPN it works so somehow the VPN is blocking the connection to the local IP of my unraid box. If I use MQTT internal to my network its fine VPN on or off.

I'm using wireguard to have all unraid outgoing connections go though a VPN. Pain to setup as some of the options needed wont allow. I had to export the config it generated, make the changes and import it back, and if I make any change though the UI it fails again. Either way, that's setup now but it seems like any outside data I forward through my router to a docker doesn't work. Example, I am using a MQTT docker with bridge networking. I setup port forward on my router for 1883 to unraid. It was working fine/is working fine. If I have the VPN connected I cant connect to the MQTT server from outside my network. If I turn off the VPN it works fine again. Is there something i need to do to make this work?

Ok so in my cron just add bash? * * * * * /boot/custom/influxdb/influxStats_1m_Disk.sh > /dev/null 2>&1 change to * * * * * bash /boot/custom/influxdb/influxStats_1m_Disk.sh > /dev/null 2>&1

So to confirm, I cant run any of my bash scripts from the flash anymore? I have a folder of scripts that execute as cronjobs. Do I need to copy these somewhere using the Go file then execute them from there? Is there a standard location to copy them to?

Hm, I had that already toggled and it wasnt showing up. I tried it on and off a few times and randomly it worked. Wierd. Thanks.

For some reason I cant figure out how to share my unassigned drive. I have it mounted I can see it in /mnt/disks but I cant figure out where to turn on smb share. The instructions on the first page say to toggle the switch but I cant find a switch for the life of me. I can see that in the settings Enable NFS Export is set to No but its greyed out so I cannot adjust it. The rest of the settings on that page I have tried all without success. Can anyone point me in the right direction?

Anyone notice extremely slow upload speed the last week or 2? Trying to figure out if its me somehow or crashplan.

I think you nailed it. So what I did is enable Grafana's auth.proxy so that it passes the username from the basic authentication to Grafana to use for logins instead of the Grafana login box. I added the header I needed to pass that to nginx conf. Made a few other changes like usernames and some settings I found online (may not have done anything but it seems to work so I left them in, and it seems to stay logged in for at least longer. Was able to do an hour before I posted this. Hopefully it continues. Thanks Aptalca for pointing me in the right direction.

It seems about 10 minutes. Give or take. There is authentication in Grafana but its not .htaccess based.

This has been bothering me for awhile but I never got to it. My Grafana docker is reverse proxied with .htaccess but it seems to timeout frequently and need to re-login. Even trying to re-login seems to not really work and I need to refresh the page then it works fine for a bit until I get the login prompt again. Has anyone else run into this issue or have a suggestion on what to try? My dockers are all accessible with a subdomain and all work fine except for this random, I'm assuming, timeout issue. This is my configuration which seems pretty standard. # Grafana # server { listen 443 ssl; root /config/www; index index.html index.htm index.php; server_name grafana.*; ssl_certificate /config/keys/letsencrypt/fullchain.pem; ssl_certificate_key /config/keys/letsencrypt/privkey.pem; ssl_dhparam /config/nginx/dhparams.pem; ssl_ciphers 'ECDHE-RSA-AES128-GCM-SHA256:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-AES256-GCM-SHA384:DHE-RSA-AES128-GCM-SHA256:DHE-DSS-AES128-G$ ssl_prefer_server_ciphers on; client_max_body_size 0; location / { auth_basic "Restricted"; auth_basic_user_file /config/nginx/.htpasswd; include /config/nginx/proxy.conf; proxy_pass http://192.168.254.3:3000; } }

Could have been a one-time issue. I restarted the docker and the error didnt come back. I was just thinking it may be a one time error then it doesnt appear again because it says it is unable to monitor in real-time so maybe it disabled that function. I'll monitor it some more and see. I bumped it up to 3.5million as i have the ram to spare and i did verify that within the docker it is the same as the host as you described. Thanks for the quick reply.

Anyone have crashplan complaining about running out of inotify watches? On my unraid system I should have 2 million available and I can confirm I have free watches through SSH on unraid. Any idea why crashplan is complaining? I assume its inside the docker that it ran out. Is there a way to increase this inside the docker?

Thanks wierd. I just checked mine and it expires after 1 year. But is that price accurate? it cant be... Product Name Payment Method Price Auto Renewal Date Expiration Date CrashPlan for Small Business Migration (Details) 2.49 USD 9/13/2018 9/17/2018

Thanks for the quick update. That fixed the issue.

Anyone else have issues with the docker. Keeps saying its not up to date but then fails to update. (My docker is up to date)

I just installed this and a bit of a newb with it but I cant seem to get it to run. I keep getting this error when i add a RTSP source: Memory map file '/dev/shm/zm.mmap.2' does not exist. zmc might not be running. I can view the source in VLC and it plays fine so I dont think that is the issue: rtsp://192.168.254.76:8554/unicast Updated the settings and now i get these errors:

Is there any way to start the docker with the old certificates and bypass the removing and attempt at reissue? I have backups but if i add them in they get distroyed when the container starts.

Great, I tried a bunch of things before seeing this thread and now i am banned for a week. All my services are now down for a week without a way around them? This is pretty crap.