Install Discourse for development using Docker

[PetzInfo (Friedrich Commichau)]

There is no way to docker Discourse on a Mac with a M1 chip currently right?

[pento (Gary)]

When running d/boot_dev --init, I was running into this error:

ActiveRecord::NoDatabaseError: We could not find your database: discourse_development. Available database configurations can be found in config/database.yml. (ActiveRecord::NoDatabaseError)

I’m not sure if this is a bug in the boot_dev script, or if the discourse_development database is supposed to be created somewhere else. Either way, I was able to fix it with this change:

diff --git a/bin/docker/boot_dev b/bin/docker/boot_dev
index 89fdcf6f65..0e9c5ce7d5 100755
--- a/bin/docker/boot_dev
+++ b/bin/docker/boot_dev
@@ -122,6 +122,7 @@ echo "pnpm install..."
 
 if [ "${initialize}" = "initialize" ]; then
     echo "Migrating database..."
+    "${SCRIPTPATH}/rake" db:create
     "${SCRIPTPATH}/rake" db:migrate
     RAILS_ENV=test "${SCRIPTPATH}/rake" db:migrate

After that, the setup process ran properly, following the instructions in the OP.

I’m not familiar enough with the Discourse codebase to know whether this is the correct fix, but I’m happy to submit a PR if that would help.

[nat (Natalie T)]

Oh yes please create a PR and link it here. Thanks!

[pento (Gary)]

Done!

DEV: Create the DB before migration in the Docker environment

[Ivan.bacher]

Any luck with getting it to run?

[bitmage (Brandon Mason)]

@martyn_thomas

Conflict. The container name "/discourse_dev" is already in use by container

I got this same error. I ran docker rm -f /discourse_dev and re-ran d/boot_dev --init and it worked this time.

[GregorSondermeier]

@Discourse

Data is persisted between invocations of the container in your source root tmp/postgres directory. If for any reason you want to reset your database run:

sudo rm -fr data

Heads up:
I’ve just noticed the folder is data/postgres instead of tmp/postgres, which is also why the “reset database” command aims to delete the data folder

[merefield (Robert)]

I thought I’d give the docker dev install another look, but keep getting:

Caused by:
PG::ConnectionBad: connection to server on socket "/var/run/postgresql/.s.PGSQL.5432" failed: No such file or directory (PG::ConnectionBad)
        Is the server running locally and accepting connections on that socket?

Yet I can confirm the service is up (and works when using my normal non-docker “manual” install of discourse)

~/discourse_docker/discourse$ sudo service postgresql status
● postgresql.service - PostgreSQL RDBMS
     Loaded: loaded (/lib/systemd/system/postgresql.service; enabled; vendor preset: enabled)
     Active: active (exited) since Mon 2025-03-10 17:26:13 GMT; 6min ago
    Process: 2052109 ExecStart=/bin/true (code=exited, status=0/SUCCESS)
   Main PID: 2052109 (code=exited, status=0/SUCCESS)

Mar 10 17:26:13 black-knight systemd[1]: Starting PostgreSQL RDBMS...
Mar 10 17:26:13 black-knight systemd[1]: Finished PostgreSQL RDBMS.

Does Docker Dev have specific and different Postgres connectivity needs?

[bitmage (Brandon Mason)]

I found the default :debug log levels to be way too verbose - the SQL queries drown out important errors. config/environments/development.rb allows for the setting of the DISCOURSE_DEV_LOG_LEVEL, but I don’t know if this can be done through the d/rails script?

Anyway I just temporarily hard coded it in config/environments/development.rb to config.log_level = :info and that solved the problem for me. Might help others as well.

[ganncamp (G Ann Campbell)]

I’m trying to spin this up with the Data Explorer plugin because I’d like to poke at the DE automation.

I got the container to start.

Then stumbled through “installing” the plugin by cloning the repo and symlinking it into the disourse/plugins directory (is that what I was supposed to do? The guide really could be a lot clearer on that point).

I enabled the plugin through the UI. But when I try to access it, I get an error.

All that shows up in the log is this:

Processing by DiscourseDataExplorer::QueryController#index as JSON
Completed 500 Internal Server Error in 62ms (ActiveRecord: 9.8ms (5 queries, 0 cached) | GC: 2.9ms)

What am I supposed to do with this? Where are the details?

[Arkshine]

@Discourse

If a db migration is needed:

d/rake db:migrate RAILS_ENV=development

Have you run the migrations?

[bitmage (Brandon Mason)]

I did not have very good luck with symlinking my plugins folder:

Can we use `.gjs` for route templates?

I found in the d/rails s logs: No such file or directory @ rb_sysopen - /src/app/assets/stylesheets/plugin-cf.scss. So even though the plugin is symlinked into the /plugins folder as county-fence, and my plugin is explicitly named county-fence in plugins.rb, some code is reading the underlying directory name plugin-cf and making assumptions about what the compiled CSS name for my plugin is. Is this a bug or a feature? I think the moral of this story is do not under any circumstances…

I ended up moving to a normal folder as a result (see my last comment in that thread).

I think there are some incompatibilities currently with Docker volumes, file watchers, and symlinks.

[ganncamp (G Ann Campbell)]

@Arkshine

Have you run the migrations?

Thanks for the reply @Arkshine! What should have alerted me that this needed doing? I did check the tutorial on installing a plugin, but I guess the plugin the tutorial uses as an example doesn’t need migrations?

In the meantime, I don’t know if this gets me over the hump because I’ve run smack into needing to get my company’s proxy cert into the Docker container & I haven’t done much with Docker before

[pfaffman (Jay Pfaffman)]

@ganncamp

I’ve run smack into needing to get my company’s proxy cert into the Docker container

What problem are you solving? A development install is generally intended to be accessible only to the localhost (i.e., your computer) where the install is running.

Or maybe you’re trying to debug something that requires a cert like an auth system?

[ganncamp (G Ann Campbell)]

When I tried to start my image back up yesterday, it wanted to download gems?

Installing gems...
Fetching source index from https://rubygems.org/

Retrying fetcher due to error (2/4): Bundler::Fetcher::CertificateFailureError Could not verify the SSL certificate for https://rubygems.org/.
There is a chance you are experiencing a man-in-the-middle attack, but most likely your system doesn't have the CA certificates needed for verification. For information about OpenSSL certificates, see https://railsapps.github.io/openssl-certificate-verify-failed.html.

Retrying fetcher due to error (3/4): Bundler::Fetcher::CertificateFailureError Could not verify the SSL certificate for https://rubygems.org/.
There is a chance you are experiencing a man-in-the-middle attack, but most likely your system doesn't have the CA certificates needed for verification. For information about OpenSSL certificates, see https://railsapps.github.io/openssl-certificate-verify-failed.html.

Retrying fetcher due to error (4/4): Bundler::Fetcher::CertificateFailureError Could not verify the SSL certificate for https://rubygems.org/.
There is a chance you are experiencing a man-in-the-middle attack, but most likely your system doesn't have the CA certificates needed for verification. For information about OpenSSL certificates, see https://railsapps.github.io/openssl-certificate-verify-failed.html.

Could not verify the SSL certificate for https://rubygems.org/.
There is a chance you are experiencing a man-in-the-middle attack, but most likely your system doesn't have
the CA certificates needed for verification. For information about OpenSSL certificates, see
https://railsapps.github.io/openssl-certificate-verify-failed.html.