You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
#25 Now that IAM seems fully supported it's looking like pytest-localstack is the best and only game in town for testing mocks in multi-account environments. And multi-account is becoming more widely used, as the GCP-originated paradigm of treating an account as a project for security isolation is growing.
From what I can tell, pytest-localstack's mapping of "docker container to session fixture to account" is a great way to go, and it would be even better if we could mock account numbers on a per-fixture basis. Currently all sessions have a mocked account number of '000000000000', but it might be simple enough to add mock_account_number as a parameter to the fixture so we could devise cross-account tests that took ARNs into account.
Automatic generation of account numbers might be possible or desirable too, but a fixture parameter that overrides the zeros might easily be enough.
Just a thought to make a good thing better ...
The text was updated successfully, but these errors were encountered:
#25 Now that IAM seems fully supported it's looking like pytest-localstack is the best and only game in town for testing mocks in multi-account environments. And multi-account is becoming more widely used, as the GCP-originated paradigm of treating an account as a project for security isolation is growing.
From what I can tell, pytest-localstack's mapping of "docker container to session fixture to account" is a great way to go, and it would be even better if we could mock account numbers on a per-fixture basis. Currently all sessions have a mocked account number of '000000000000', but it might be simple enough to add mock_account_number as a parameter to the fixture so we could devise cross-account tests that took ARNs into account.
Automatic generation of account numbers might be possible or desirable too, but a fixture parameter that overrides the zeros might easily be enough.
Just a thought to make a good thing better ...
The text was updated successfully, but these errors were encountered: