assumption: with a larger amount of incoming data, it will take longer to find errors and notify the user
test: this will actually benchmark the application accessing our provenance information.
To save money for leasing cloud resources we could use data from a smaller test deployment, multiply it and send it within the same time to the DB
Metrics: time until we actually find an error and push a message to the user, amount of input data per time.
assumption: with a larger amount of incoming data, it will take longer to find errors and notify the user
test: this will actually benchmark the application accessing our provenance information. To save money for leasing cloud resources we could use data from a smaller test deployment, multiply it and send it within the same time to the DB
Metrics: time until we actually find an error and push a message to the user, amount of input data per time.