数据流错误 - “IOException：无法写入GCS路径..”“后端错误500”

One of our pipelines threw the below error. First time we've seen it. We're running over around 625 million rows from a BigQuery table. The job still completed, and was recorded as "successful" in the console. But we are concerned that maybe the file that Dataflow couldn't write to GCS (Dataflow writes to GCS to then load into BigQuery) was not loaded into BigQuery, and thus we are now missing some data.One of our pipelines threw the below error. Fir