Ruby on Rails

Rails has built-in multi-database support since version 6.

Read replica routing

Point both roles at halephant with different users in config/database.yml:

config/database.yml

production:
  primary:
    host: halephant
    port: 6432
    database: myapp
    username: myapp
    password: ...
  primary_replica:
    host: halephant
    port: 6432
    database: myapp
    username: myapp_ro
    password: ...
    replica: true

Enable automatic connection switching in config/application.rb:

config/application.rb

config.active_record.database_selector = { delay: 2.seconds }
config.active_record.database_resolver =
  ActiveRecord::Middleware::DatabaseSelector::Resolver
config.active_record.database_resolver_context =
  ActiveRecord::Middleware::DatabaseSelector::Resolver::Session

Rails routes GET and HEAD requests to the replica after a configurable delay (default 2 seconds after the last write). POST, PUT, DELETE, and PATCH requests use the primary.

For manual control:

app/models/concerns/read_routing.rb

ActiveRecord::Base.connected_to(role: :reading) do
  # Queries go to the replica.
end

Rails doesn't send BEGIN READ ONLY. The separate user approach lets halephant handle routing. See the read replica guide for halephant configuration.