The steps of installing the application are not completely clear

[YakovMyagkov]

Hello! I was very interested in your project, I tried to assemble it on my site, but I could not assemble the application from the source code, could you make a more detailed instruction (step by step) how to do it?
I have absolutely no experience in erlang.
It was possible to make an instance through Docker, but I would also like to fully deploy it on the machine as a service

[dgulinobw]

Unfortunately, the full Ansible based deployment we use cannot be Open Sourced.

You can run the central portion of dog in Docker - you can expose those services external to the machine via Docker - then all you have to deploy is the individual dog_agents. In fact, I think that will be the recommended way of doing things.

A better full deployment HOWTO is still on the TODO.

[YakovMyagkov]

Yes, installing docker is easy in terms of a central server, but there are problems with installing the agent, as I wrote earlier in another topic. As I understand correctly, are you offering to deploy the agent on another machine, too, as a docker image, or is it still a full-fledged installation as a service?

[dgulinobw]

1. The delay should be 10-15 seconds tops; more than that and it's not going to ever work.
   Did you try iptables-legacy-save instead of iptables-save?
2. The nginx routing in that docker image requires a hostname to route requests to the proper backend. You'll need a DNS entry for a central dog server - use that as REACT_APP_DOG_API_HOST=http://yourdog.yourdomain.com
   see: https://hub.docker.com/r/jwilder/nginx-proxy

[YakovMyagkov]

I used 2 commands, my result looks like:

[YakovMyagkov]

Profile in Dog_trainer it looks like this: