Django Json Mirror

Mirror and sync Django models in a json in another database.

Install

Clone then:

pip install -r django-jsonmirror/requirements.txt

Add "jsonmirror", to installed apps.

Settings

Declare the databases:

SITE_SLUG = "my_site"
MIRROR_DATABASES = {
  "main": {
      "name": SITE_SLUG,
      "type": "sqlite"
  },
  "alt": {
      "name": SITE_SLUG,
      "type": "rethinkdb"
  }
}

The SITE_SLUG setting is required

Register models

Declare the models you want to mirror in settings.py and set the database options for each model to mirror:

MIRRORED_MODELS = {
  'django.contrib.auth.models.User': {
      "alt": {
         "table": "users",
         "imutable": False
      }
  },
  'myapp.models.MyModel': {
       "main": {
          "table": "mytable",
          "imutable": False,
          "soft_delete": False
      },
      "alt": {
          "table": "pages"
      }
  }
}

Note that table is required

Database options

Imutability

Set to True by default: the database is write only

Soft delete

When imutability is turned of the soft delete option, enabled by default, will mark documents as deleted. Set it to False to delete the documents in the database

Backends

Rethinkdb

Settings:

# required
MIRROR_DB = "mydb"
MIRROR_TABLE = "mymirrortable"

# optional:
RETHINKDB_HOST = "127.0.0.1" # default
RETHINKDB_USER = "admin" # default is None
RETHINKDB_PASSWORD = "mypassword" # default is None

SqlAlchemy

Settings:

# required
MIRROR_SQL_ENGINE = 'sqlite:///mirror.db'

Refer to SqlAlchemy docs for how to setup db urls

Note: the sql backend is limited for now: it saves the data to an unique table named mirror with django model path and the instance pk to use as compound index to retrieve the data.

Models synchronization

The models declared for mirroring will be auto synchronized in the secondary databases at each save/delete operation performed on them

The command python manage.py sync_models will mirror all the models declared in the MIRRORED_MODELS setting. They will be auto created/updated in the secondary databases using this command.

Todo

• SqlAlchemy backend
• Redis backend
• Support DRF serializers
• Support custom serializers
• Better handling of connection failure