I use Hasura for getting an automatically generated CRUD-layer to the database in Triglog. It works really well, and has saved me a lot of time writing boring and error-prone boilerplate code. The only thing that has slowed me down some is jumping into it not really knowing GraphQL beyond the very basics of making a query.

These are the basic steps needed to deploy Hasura if you’re using Dokku:

From https://github.com/hasura/graphql-engine-heroku/issues/36#issuecomment-609016274

dokku apps:create hasura
dokku domains:add hasura hasura.my.domain.com
dokku postgres:create hasura-db
dokku postgres:link hasura-db hasura
dokku config:set hasura HASURA_GRAPHQL_DATABASE_URL="<info from above command>" HASURA_GRAPHQL_ADMIN_SECRET="blahblah" HASURA_GRAPHQL_ENABLE_CONSOLE="true" 
dokku proxy:ports-set hasura http:80:8080
docker pull hasura/graphql-engine
docker tag hasura/graphql-engine dokku/hasura
dokku tags:deploy hasura
dokku letsencrypt hasura