appengine set up local host with datastore for testing

Question

I have tried to follow googles documentation on how to set up local development using a database ( https://cloud.google.com/appengine/docs/standard/python/tools/using-local-server#Python_Using_the_Datastore ). However, i do not have the experience level to follow along. I am not even sure if that was the right guide. The application is a Django project that uses python 2.7. To run the local host, i usually type dev_appserver.py --host 127.0.0.1 .

My questions are:

1. how do i download the data store database on google cloud. I do not want to download the entire database, just enough data to populate local host so i can do tests

2. once the database is download, what do i need to do to connect it to the localhost? Do i have to change a parameter somewhere?

3. do i need to download the datastore? Can i just make a duplicate on the cloud and then connect to that datastore?

4. When i run localhost, should it not already be connected to the datastore? Since the site works when it is running on the cloud. Where can i find the connection URI?

Thanks for the help

Answer 1

The development server is meant to simulate the whole App Engine Environment, if you examine the output of the dev_appserver.py command you'll see something like Starting Cloud Datastore emulator at: http://localhost:PORT . Your code will interact with that bundled Datastore automatically, pushing and retrieving data according to the code you wrote. Your data will be saved on a file in local storage and will persist across different runs of the development server unless it's explicitly deleted.

This option doesn't provide facilities to import data from your existing Cloud Datastore instance although it's a ready to go solution if your testing procedures can afford populating the local database with mock data through the use of a custom created script that does so programmatically. If you decide for this approach just write the data creation script and execute it before running the tests.

Now, there is another option to simulate local Datastore using the Cloud SDK that comes with handy features for your purposes. You can find the available information for it under Running the Datastore Emulator documentation page. This emulator has support to import entities downloaded from your production Cloud Datastore as well as for exporting them into files.

Back to your questions:

1. Export data from the Cloud instance into a GCS bucket following this , then download the data from the bucket to your filesystem following this , finally import the data into the emulator with the command shown here .
2. To use the emulator you need to first run gcloud beta emulators datastore start in a Cloud Shell and then in a separate tab run dev_appserver.py --support_datastore_emulator=true --datastore_emulator_port=8081 app.yaml .
3. The development server uses one of the two aforementioned emulators, in both cases it is not connected to your Cloud Datastore. You might create another project aimed for development purposes with a copy of your database and deploy your application there so you don't use the emulator at all.
4. Requests at datastore are made trough the endpoint https://datastore.googleapis.com/v1/projects/project-id although this is not related to how the emulators manage the connections in your local server.

Hope this helps.