New instalation with doodba-copier-template in production error "Internal Server Error"

[JorgeQuinteros]

Hello,
I have followed the steps to install an Odoo instance with dooba-copier-template:
https://github.com/Tecnativa/doodba-copier-template
https://github.com/Tecnativa/doodba-copier-template/blob/main/docs/daily-usage.md#production
But I am getting the following error:

The log showed that I had to initialize the base, so I launched the commands recommended here but for production.
https://github.com/Tecnativa/doodba-copier-template/blob/main/docs/faq.md#why-is-odoo-saying-that-its-database-is-not-initialized
However this did not work, although I was no longer getting errors in the odoo log.

I started to check the log of the other containers, and in the inverseproxy_odoo_1 container I got the following:

It definitely doesn't seem to be finding the base module so I updated the repositories to verify that it would download Odoo OCB, it seems that it does, however the same error kept coming up.

In the folder structure I don't see that Odoo OCB has been downloaded, if it was for permissions it would give error when doing buil --pull, right?
Do you know if there is any extra step to do that is not in the documentation? any help would be appreciated.

Thanks,

[joao-p-marques]

Hi @JorgeQuinteros

This seems like an error caused by having multiple databases in an instance, and when Odoo tries to access them, looks like at least one of them (prod) is not correctly initialized.

The first thing that might help you understand what is happening is taking a look at the different environments we use with Doodba Copier Template: We have a devel, a test, and a prod environment, each one corresponding to a different docker compose file. Each of the files will then be symlinked to docker-compose.yaml, so that you can interact with the containers without having to specify the file name each time. Taking a look at each of the files, you can also see that, depending on the environment, Odoo will access different databases. From what you said in the comment, I guess you are using the production one.

OTOH, the invoke tasks we have and that are mentioned in the docs are targeted at the development workflow, and they will sometimes change the symlink from docker-compose.yaml back to devel.yaml. So, if you are using those, you might be initializing one DB and then connecting to another from Odoo.

I also don’t understand what you mean when you mention the logs from 2 different Odoo containers, one showing everything correct, apparently, and the other saying that the prod DB was not correctly initialized. However, there should only be one Odoo container, so maybe check your running containers and docker compose files to see if there is something wrong there.

I also have some doubts from seeing that you have another file (inverseproxy.yaml) in you project folder. It is true that this project relies on an inverse proxy (like Traefik) that should be created externally, but I wonder if some of its configurations might be causing the problem here. Can you provide its contents and explain a little more what you are doing command-wise?

Finally, you mention there might be an issue aggregating the repositories. Indeed, it is weird that no code appears in odoo/ custom/src. However, if you are seeing the logs you mention, then the Odoo code must be running somewhere, either in your machine, or coming from a pre-built image that includes that code.

[JorgeQuinteros]

Hello @joao-p-marques
First of all, thank you very much for taking the time to read and answer.

I did not use the invoke command precisely because I knew it was used for the development environment.

Regarding the two odoo containers, I am equally confused, probably I have made a mistake somewhere or maybe it is by wrongly launching the command to start the inverseproxy.yaml, as you said.
This is the content of the inverseproxy.yaml file(I got the content from here: https://github.com/Tecnativa/doodba-copier-template/blob/main/docs/faq.md#how-to-bootstrap-the-global-inverse-proxy):
And I execute it with: docker-compose -f prod.yaml -p inverseproxy -f inverseproxy.yaml up -d

version: "2.1"

services:
  proxy:
    image: traefik:1.6-alpine
    networks:
      shared:
      private:
      public:
    volumes:
      - acme:/etc/traefik/acme:rw,Z
    ports:
      - "80:80"
      - "443:443"
    depends_on:
      - dockersocket
    restart: unless-stopped
    privileged: true
    tty: true
    command:
      - --ACME.ACMELogging
      - --ACME.Email=admin@nextodoo.com
      - --ACME.EntryPoint=https
      - --ACME.HTTPChallenge.entryPoint=http
      - --ACME.OnHostRule
      - --ACME.Storage=/etc/traefik/acme/acme.json
      - --DefaultEntryPoints=http,https
      - --EntryPoints=Name:http Address::80 Redirect.EntryPoint:https
      - --EntryPoints=Name:https Address::443 TLS
      - --LogLevel=INFO
      - --Docker
      - --Docker.EndPoint=http://dockersocket:2375
      - --Docker.ExposedByDefault=false
      - --Docker.Watch

  dockersocket:
    image: tecnativa/docker-socket-proxy
    privileged: true
    networks:
      private:
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
    environment:
      CONTAINERS: 1
      NETWORKS: 1
      SERVICES: 1
      SWARM: 1
      TASKS: 1
    restart: unless-stopped

networks:
  shared:
    internal: true
    driver_opts:
      encrypted: 1

  private:
    internal: true
    driver_opts:
      encrypted: 1

  public:

volumes:
  acme:

PD:

I have checked the containers and I have clearly seen that I have mistakenly created an extra Odoo container. Probably because I lack more experience with docker and docker-compose
I stopped it the extra odoo container and restarted with docker-compose restart and it worked correctly!

Thank you very much again for responding, and we apologize for the inconvenience.

[pleirb]

@JorgeQuinteros How are you? I have followed the same steps to try to work in production, but I get a "Bad Gateway" in the browser

I use docker-compose -f prod.yaml up -d to try to run, and I don't know if there is something else to run or just this is enough
I have also used the command docker-compose -p inverseproxy -f inverseproxy.yaml up -d with the corresponding file that is mentioned in the documentation.
This is my output from docker ps

And this is my way out docker network ls

Then, to test, since I don't have the domain yet, I've edited the /etc/hosts file and added the domain I'm testing in my browser like this: 127.0.0.1 web.host.com
Many thanks in advance to anyone who dares to give me guidance.

[pleirb]

@JorgeQuinteros¿Cómo estás? He seguido los mismos pasos para intentar trabajar en producción, pero obtengo un "Bad Gateway" en el navegador que uso para intentar ejecutar, y no sé si hay algo más para ejecutar o simplemente esto es suficiente . también he usado el comando con el archivo correspondiente que se menciona en la documentación. Esta es mi salida de docker ps Y esta es mi salida docker network ls Luego, para probar, dado que aún no tengo el dominio, edité el archivo /etc/hosts y agregué el dominio en el que estoy probando mi navegador así: 127.0.0.1 web.host.com Muchas gracias de antemano a quien se atreva a orientarme. docker-compose -f prod.yaml up -d``docker-compose -p inverseproxy -f inverseproxy.yaml up -d

Finally, to solve, what they comment here has helped me: #418