dbt-labs / dbt-labs/dbt-adapters
[Bug] 🐞 dbt-bigquery does not retry on 503 error
- Dominant language
- Python
- Stars
- 233
- Forks
- 362
- Avg merge
- 3d 22h
- Merged PRs (30d)
- 9
Description
### Summary
We intermittently get 503 errors from GCP/BigQuery, which result in job failures. These 503 do not get logged as [backend errors](https://cloud.google.com/bigquery/docs/error-messages). On the contrary, I see the job successfully finished in 0 seconds. The following shows up in "Bigquery Job explorer" section. I believe that is because this is **Runtime Error** as opposed to **Backend Error**.
```
Duration: 0 sec
Bytes processed: 0 B
Bytes billed: 0 B
Slot milliseconds: 0
```
### Logs:
```
1 of 4 ERROR creating sql view model myproject.my_dataset.my_table [ERROR in 1.36s]
BigQuery adapter: Unhandled error
Runtime Error in model my_model (dbt/models/some.sql)
503 GET https://bigquery.googleapis.com/bigquery/v2/projects/my-project/jobs/ffffffff-8d8d-4747-acac-eeeeeeeeeeee?location=US&prettyPrint=false:
Visibility check was unavailable. Please retry the request and contact support if the problem persists
```
### DBT Environment
```
dbt-core: 1.9.1
dbt-bigquery: 1.9.1
OS: debian docker image
Python: v3.10
```
### Expected Behavior
dbt retries on 503 when job_retries is set to 1
### Actual Behavior
dbt errors out without retrying even when job_retries is set to 1
Contributor guide
Assessment
This issue has not been assessed yet.