ingestr v1.1.71 landed on October 19, 2026, and the change that matters most to operators is a BigQuery fix: a DropFirst table is now recreated when its schema no longer matches the source. Custom endpoint support for Athena and Snowflake and a Jira auth option round it out.
The full release notes and downloads are on the GitHub release page. Four changes shipped.
BigQuery: recreate DropFirst tables when the schema drifts ¶
PR #1327 fixes a long standing annoyance in the bigquery destination. When you use the DropFirst ingestion mode, ingestr drops and recreates the target table. If the source schema changed since the last run, for example a column moved from INT64 to FLOAT64, the recreated table could conflict with the new data and the load would fail.
The fix recreates the table when its schema no longer fits the incoming data. Win: pipelines with evolving source schemas stop failing on schema mismatch. Cost: a dropped table loses its partitioning and clustering configuration, so those are rebuilt from the current source definition, not your old DDL. If you rely on hand tuned table settings under DropFirst, verify them after upgrading.
Custom endpoints for Athena and Snowflake ¶
PR #1325 adds custom endpoint configuration for both Athena and Snowflake. This matters if you route traffic through a private endpoint, a VPC endpoint service, or an internal proxy instead of reaching the default regional host directly.
Athena and Snowflake are also sources ingestr commonly pulls from inside data platform networks where the public service hostname is not reachable. Custom endpoints remove the need for DNS hacks or sidecar proxies just to satisfy a fixed hostname.
Jira service account tokens ¶
PR #1329 adds support for Jira service account tokens via the Atlassian API gateway. Until now, a Jira ingestion needed a user style credential, which is awkward for headless pipelines and clashes with how Atlassian is steering API access for automated callers.
If your Jira source runs on a schedule, this lets you use a dedicated service account identity instead of a personal token. That is the better pattern for audit trails and for rotating credentials without touching a human account.
CI speedups ¶
PR #1328 speeds up the StarRocks and MSSQL CDC integration tests. This is internal work; it does not change runtime behavior. The practical effect is faster feedback on the projects own CI and less queue time when you submit a patch.
Upgrade notes ¶
No breaking changes were flagged in the notes. Full changelog is at v1.1.70…v1.1.71. Two things to check when you upgrade:
- If you use
bigquerywithDropFirst, confirm recreated tables keep the partition and cluster settings you expect. - Update any Jira ingestion to a service account token if your organization is phasing out personal API tokens.
Where to get it ¶
- Release page: ingestr v1.1.71
- Repository: bruin-data/ingestr
- Tag:
v1.1.71