Environment details

• Python version: 3.11.10
• pip version: 24.2
• google-cloud-storage version: 2.18.2

Description

For resumable media calls in cloud storage (ex: download_as_string), the provided google.api.core.Retry is converted to a google.resumable_media.RetryStrategy. However, the meaning of the provided deadline appears to differ between these two classes.

For google.api.core.Retry, deadline seems to be defined as "the maximum amount of time a function can block"

However, for google.resumable_media.RetryStrategy, max_cumulative_retry (what deadline corresponds to) is defined as "The maximum total amount of time to sleep during the retry process."

There appears to be a mismatch between the definitions though. One refers to the total time the function blocks while the other refers to the cumulative time spent sleeping between retries.

If this is indeed a mismatch is it possible to update the documentation of resumable media operations to clarify that they have a different retry strategy?