-
Notifications
You must be signed in to change notification settings - Fork 35
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Installation failed on Ubuntu 18.04 and Ubuntu 20.04 #66
Comments
I've used this script successfully before, for Decidim 19.0, but I came back to do a new installation and found that it breaks with exactly the same behaviour described above, an ERR_CONNECTION_RESET when the http request is done using the domain name. I should add that the script runs without any problems and indicates "Servers installed successfully" at the end. I also tried doing the installation manually, following the step by step instructions, and ended up with the same result. |
Solved it. It does appear that SSL is now forced, yet the script doesn't enable it on the NGINX server. So you just have to configure SSL following a standard process: https://www.techrepublic.com/article/how-to-enable-ssl-on-nginx/ The only decidim-specific modifications are to append the four lines relevant to SSL configuration into decidim.conf, and that the symbolic linking between /sites-available/ and /sites-enabled/ needs to be reversed. |
Spoke a little too quickly. The above fix works well up until the mail configuration, at which point "ActionMailer" overflows, an error that seems to be related to this problem and which prevents emails from being sent via smtp. |
Could you post the full output of the script? this way is much easier to know where was the failure |
Okay, I'll rerun it and capture it... although, if I understand correctly, the script doesn't touch SSL, yet configuring SSL seemed to fix the problem? |
sorry, i though the problem was with the installation, so, if I understand i correctly, the installation finished just fine but you can't reach the server? |
And, it's true that, since version 0.22, ssl is forced by default, you can change that by editing the file config.force_ssl = false |
Okay, that might be all we need! Configuring SSL made the server reachable, but there was still some kind of conflict with sending emails via smtp. Hopefully this parameter will resolve both. I'll try in the morning and report back here. |
SMTP should be independent, try to configure SMTP in the |
Yes, I try to stick with the /system admin. The interesting thing was that while the error in the log was obscure (buffer overflow), Stackoverflow indicated that the problem is very often when an SSL message is sent out without SSL being flagged as a parameter (e.g. when https is being used, but ssl = true is not included in the parameters for the message formatting), which seemed related. Will test it tomorrow and see. |
Still having trouble getting smtp to work with mailgun.org. Any idea what this could be? E, [2020-11-19T12:23:02.109089 #3055] ERROR -- : [ActiveJob] [ActionMailer::DeliveryJob] [6dcf4cbf-b739-46e7-9748-142af3f86300] Error performing ActionMailer::DeliveryJob (Job ID: 6dcf4cbf-b739-46e7-9748-142af3f86300) from Async(mailers) in 30420.01ms: Net::OpenTimeout (execution expired): I see in the notes that it's often IPv6 routability that's the issue, but it's still a problem after IPv6 is disabled. |
I solved the above connectivity problem, and now have everything pointed at gmail.com, but we're now at an "end of file" error. Going to explore and see if I can find a solution. E, [2020-11-19T14:15:22.999145 #11039] ERROR -- : [ActiveJob] [ActionMailer::DeliveryJob] [46644e29-bfa6-4c7b-964e-5c1c978e67df] Error performing ActionMailer::DeliveryJob (Job ID: 46644e29-bfa6-4c7b-964e-5c1c978e67df) from Async(mailers) in 11046.58ms: EOFError (end of file reached): |
Got it working. Used https://www.smtper.net/ to verify the SMTP server access was correct, then narrowed it down to closed ports on the Digital Ocean server. Thanks for the help! |
great to know, hopefully your debugging may be helpful to other people! |
hello, |
One interesting tip, SSL becomes incredibly simple to install in Ubuntu 20.04 with the certbot snap. It's just two commands.. https://tecadmin.net/how-to-setup-lets-encrypt-on-ubuntu-20-04/ I'll let the experts work on your '403' issue. |
I undersant that you are using either the config from There's missing parameters in you server configuration with ssl, it should look like this: server {
root /home/decidim/decidim-app/public;
index index.html index.htm index.nginx-debian.html;
server_name connect-citoyen.fr www.connect-citoyen.fr;
client_max_body_size 32M;
passenger_enabled on;
passenger_ruby /home/decidim/.rbenv/shims/ruby;
rails_env production;
listen [::]:443 ssl ipv6only=on; # managed by Certbot
listen 443 ssl; # managed by Certbot
ssl_certificate /etc/letsencrypt/live/connect-citoyen.fr/fullchain.pem; # managed by Certbot
ssl_certificate_key /etc/letsencrypt/live/connect-citoyen.fr/privkey.pem; # managed by Certbot
include /etc/letsencrypt/options-ssl-nginx.conf; # managed by Certbot
ssl_dhparam /etc/letsencrypt/ssl-dhparams.pem; # managed by Certbot
}
... |
yes ! it works ! Thanks |
I ve followed many times the 2 installation guide for Ubuntu 18.04 and 20.04 on clean VPS. The result is always the same : by default the domain name link to a https connection and the site is unreachable (ERR_CONNECTION_RESET), if i type directly a non secure link to my domain with http, it leads to nginx welcome page... It seems to be the same problem mentioned in topic #45 or #63...
Any help ?
The text was updated successfully, but these errors were encountered: