JBodkin created BEAM-9828:
-----------------------------
Summary: BigQueryIO attempts to repeat the same job id for batch
load when the table does not exist
Key: BEAM-9828
URL: https://issues.apache.org/jira/browse/BEAM-9828
Project: Beam
Issue Type: Bug
Components: io-java-gcp
Affects Versions: 2.19.0
Reporter: JBodkin
I've noticed that BigQueryIO attempts to repeat the same job id for batch
loading when the table does not exist. I believe this happens due to the check
in `getJob` that looks at the status output.
[https://github.com/apache/beam/blob/master/sdks/java/io/google-cloud-platform/src/main/java/org/apache/beam/sdk/io/gcp/bigquery/BigQueryServicesImpl.java#L361]
The output from the GCP API for get job has the following:
{code:java}
"status": {
"errorResult": {
"message": "Not found: Table xxx was not found in location europe-west2",
"reason": "notFound"
},
"errors": [
{
"message": "Not found: Table xxx was not found in location
europe-west2",
"reason": "notFound"
}
],
"state": "DONE"
},
{code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
