Host Your Django Project on Heroku (Summary)

Thank you for this course. I was successful without Google’s help at all. Give me the direction, please:

1. How to move data from local sqlite of the project to pg in heroku
2. How to make a deal with static files of models. The idea was to add objects from admin
3. How to create superuser of the project in heroku

It seems I’m waiting for the part 2 of the course.

@allapopovarus Thanks, and I’m glad that you enjoyed this course 😊

1) How to move data from local sqlite of the project to pg in Heroku?

In general, you should let Django’s ORM handle your data management through migrations to keep everything consistent. Other than that, SQLite uses a different SQL dialect than PostgreSQL, so you won’t be able to dump your schema in one database management system and import it directly into another:

$ sqlite3 db.sqlite3 .dump > dumpfile.sql

However, if you really insist and know what you’re doing, then you can try using Django’s dumpdata and loaddata commands, which use JSON for serialization by default. First, run dumpdata locally, assuming you’re local Django settings are still hooked up to SQLite:

(venv) $ python manage.py dumpdata > data.json

This will save the contents of your local database as a fixture to a local file named data.json. Next, you’ll need to start a one-off dyno on Heroku to take advantage of the remote database’s settings:

(venv) $ heroku run bash

Now, you need somehow to transfer your local data.json file into that dyno. There’s no straightforward way as far as I know, but you can use something like GitHub Gist as a proxy and then download your file with the curl or wget command:

~ $ wget https://gist.githubusercontent.com/(...)/data.json
~ $ python manage.py loaddata data.json

Afterwards, you can exit the dyno by hitting [Ctrl] + [C]. If everything goes fine, then you should’ve successfully moved your data from SQLite to Heroku’s PostgreSQL database.

2) How to make a deal with static files of models? The idea was to add objects from admin.

Heroku has an ephemeral or temporary file system, so you can’t save any files directly on the running containers. Instead, you should use an external service such as Amazon S3 and only store the remote file references in your relational database.

3) How to create superuser of the project in Heroku?

As long as you’ve created the tables in Heroku’s database by running the first migration either manually or through the release process in your Procfile, you can run the createsuperuser command remotely in the following way:

$ heroku run python manage.py createsuperuser

Thanks, for the answers. In fact, I found solution to upload files from admin using FileField. Which uploads files to static folder of app. It works.

@allapopovarus You’re welcome. I wonder if the static files would still be available after a dyno restart or rescaling. Have you tried restarting your web process with heroku restart?

You are right… unfortunately. So let’s try with Amazon S3.

Hello,

I thought that heroku was free but that isn’t anymore. I created an account but I was to fast now I have to deleted again.

@chris.gerrist Unfortunately, Heroku has discontinued their free plan, which is why we’ve added a side note about the new policy in the overview at the beginning of this course.