The current smoke test creates a batch to be reused for a few hours, repetitively uploads 5 mb file, and then attempts to download it from the network. In essence, only immediate uploading and downloading is tested.
What is required now is to test data availability across long spans of time.
Tasks:
[ ] manually create a batch with a large TTL (eg: 2 years)
[ ] manually upload a large amount of data (eg: 1 GB), potentially multiple files
[ ] make sure that the upload files are pinned (maybe to dev-bee-gateway nodes) so that in the case that there is a data loss, we can easily use the stewardship feature to reupload the data.
[ ] create a beekeper check that takes a list of references as input and attemps to download the references based on some timer (eg: 6 hours) with proper logging and metrics.
The current smoke test creates a batch to be reused for a few hours, repetitively uploads 5 mb file, and then attempts to download it from the network. In essence, only immediate uploading and downloading is tested.
What is required now is to test data availability across long spans of time.
Tasks:
metrics: