1609546 - spark-bigquery connector fails with 'INTERNAL: request failed: internal error' on payload_bytes datasets

Status: RESOLVED FIXED

(Data Platform and Tools :: General, defect, P1)

Description

[Anthony Miyaguchi [:amiyaguchi]]

The spark-bigquery connector uses the BigQuery Storage API to access tables within Spark. This is currently being used for mozaggregator and the prio-processor jobs. This is being tracked in the Google Cloud issue tracker here: https://issuetracker.google.com/issues/147113808

Comment 1

[Mark Reid [:mreid]]

Looks like the fix should roll out this week.

Comment 2

[Mark Reid [:mreid]]

From the linked issue, the fix went out on Feb 6 - Anthony are you still seeing the same behaviour?

Comment 3

[Anthony Miyaguchi [:amiyaguchi]]

As of 2020-02-11, I had still saw the same behavior as reported in case 21629086 in the airflow-dataproc-prod project. I received a followup on 2020-02-14 that another fix would be rolled out and announced on the public issue page, which has not happened yet.

Comment 4

[Mark Reid [:mreid]]

From the linked issue, this seems to be fixed now. We talked about this last week, and IIRC there's another related problem now - can you add an update here?

Comment 5

[Anthony Miyaguchi [:amiyaguchi]]

This has been resolved as of 2020-03-02.