madaroda

Latest update to swag appears to have blocked secure access to all my installed dockers. I can access the dockers locally via http://IP address:port, but https://anydocker.mydomain.com asks for a user name and password, none of which are accepted and result in a "403 forbidden" error. Thinking it may be related to a previous htpasswd configuration, I reinstalled swag (copying over only the proxy-conf files I am using and the entire dns-conf directory to the new installation). Still no access. I don't have an .htaccess file in config/nginx. Where do I being to look for a fix? I am stumped.

I've set up Letsencrypt/nginx wildcard on an unraid docker with my own domain. Letsencrypt certificates appear to download fine. I've got sonarr, radarr and a few more subdomains edited and renamed in proxy-confs. Letsencrypt docker opens and ends in "Server ready." Domain.me and anything.domain.me resolve to correct IP. But nothing opens. https://sonarr.domain.me et all are unable to establish a connection to the server, even though sonarr.domain.me pings correctly (to the WAN address). I suspect a config issue, but where?

Terrific work, djoss. I predict this will soon be the go-to certificate manager in unraid. Questions: How would a wildcard certificate be assembled through the proxy manager? How would we go about making the LetsEncrypt cert self-renewing? Edit: Just found that Lets Encrypt wild cards don't work yet. Hope that comes soon.

I moved all plugins to a temporary folder and rebooted. The "/" folder has the following permissions. So it does not seem as if a plugin is masking "/" incorrectly. I will check containers now. Any other ideas?

Here's the zip. tower-diagnostics-20180925-1626.zip

I'd be happy to. I downloaded the Diagnostics from the Tools page, and it is a heavily populated zip file. Which file(s) should I include?

I found the problem. It appears that somewhere along the upgrade process, the permissions for the root folder ("/") were being reset to 700 on every reboot. Adding "chmod 755 /" to the bottom of the /boot/config/file fixed the problem.

I'll keep updating in hopes someone runs into this: Via SSH, I made DOCKER_ENABLED="no" in /boot/config/docker.cfg; and rebooted. Still getting the 500 error on the webgui. The errors reported in /var/log/ngnix/error.log have changed slightly: The very last part (referrer: "http://10.10.10.111/Docker") is new, and came up AFTER I disabled docker altogether. How is that possible? Also I continue to get the following email from my unraid install after reboot: Could this be related? Do I need to add a user "nobody"? Gracias.

UPDATE: I got SSH and webgui up again by reverting to the last saved version of my flash drive (6.3.x). I upgraded to the current version (6.5.x) and SSH continues to run. The WEBGUI is down again, though. I enter my user name and password, and the browser (any of them) now goes to a "500 Internal Server Error --- nginx" page. The content of /var/log/ngnix /error.log is now different, reading: Permissions of /etc/nginx/htpasswd are at 666. Fix?

Disabled dockers as you suggested. Trying webgui through IP address now reports a "502 Bad Gateway - ngnix" page. Still no SSH access. Entering through telnet, here is the contents of /var/log/ngnix /error.log. Also, got this via email: Thanks.

A couple of days ago i added letsencrypt/nginx and organizr2 dockers. I am guessing this was the start of the problems I now have. All of my docker apps are accessible via both internal IP and external https (my own domain). However: I have no access to the unraid webgui. Log files pointed to a permissions error on .htpassword. In following some instructions on fixing that, I rebooted the box. Now I don't even have SSH access with root ("Permission denied" on entering password). I deleted all the password files in /boot/config and rebooted. Only way in is through telnet. Again, all apps work, I just have no (easy) way to configure the system. How do I get my webgui and SSH access back with only telnet at my disposal?

I figured out how to stop a container in the CLI. Stopping Letsencrypt did not fix the issue. I assumed they were related because it happened right after the LE installation. So, here I am, most likely in the wrong thread, and no solution.

Using internal IP will not work (gets the 500 error). Trying hostname (https://unraid.mydomain.com), which in DNS is a C record to mydomain.com (IP set via DDNS on router) just gives me the NGINX "Welcome to our server" page.

I configured the Let's Encrypt (nginx) docker with nzbget, organizr, sonarr, radarr, etc. dockers using my own domain. I can access all of those apps using https (https://nzbget.mydomain.com). But . . . I have lost access to the Unraid GUI. I get a 500 error on the GUI, and a "/etc/nginx/htpasswd" failed (13: Permission denied)" error in /var/log/nginx/error.log. I am trying to enter the GUI locally using http. I can SS=H into unraid, and all the docker apps are accessible. Just not the front end. Advice?

Thanks. It’s what I hoped for. Now off to buy and build.