denis256 notes and projects

ingestr v1.0.61 was published on July 1, 2026, as a focused MySQL connector release. It is not marked as a prerelease, and the main change is the split of Vitess and PlanetScale into dedicated URI schemes, with a PlanetScale gRPC receive limit fix for larger responses.

The full release notes and downloads are on the GitHub release page.

Dedicated URI schemes for Vitess and PlanetScale

The core change in v1.0.61 is that ingestr now treats Vitess and PlanetScale as separate URI schemes. The release notes call this out as mysql: split Vitess and PlanetScale into dedicated URI schemes, which is the kind of connector detail that matters during setup and incident review.

Before this release, both systems lived closer to the generic MySQL path. That can work, but it also leaves room for confusing config and unclear errors when a source behaves like MySQL but still has service specific rules. Dedicated URI schemes make the intent clearer at the connection string level.

For teams running both Vitess and PlanetScale, this is the practical part: source definitions should be easier to read. A pipeline config that names the actual backend is less ambiguous than one that asks the reader to infer the backend from driver behavior or a nearby note.

PlanetScale gRPC receives larger messages

The second operator visible fix is for PlanetScale reads through psdbconnect. ingestr now raises the gRPC receive limit above 4MB. That matters when a result or metadata response crosses the old cap and the load fails before useful work can continue.

The release notes do not give a new numeric limit, so do not treat this as a broad promise about every large PlanetScale workload. The useful signal is narrower and more concrete: the hard 4MB receive ceiling was too low for some paths, and this release lifts it.

This is most relevant for jobs that were otherwise configured correctly but hit message size errors. It is not a schema change. It is a transport limit fix in the PlanetScale path.

Cleaner MySQL fail fast messages and tests

There is also a small but useful cleanup around MySQL fail fast messages. The notes say ingestr drops a hardcoded source scheme from fail-fast messages. In plain terms, errors should be less tied to the wrong or generic scheme when Vitess and PlanetScale now have their own names.

That kind of change is easy to ignore until an on call engineer has to read the error at 02:00. Better source labels reduce the time spent proving that the config points at the backend you think it points at.

The maintainers also updated destination scheme tests for Vitess and PlanetScale. This is mostly internal test coverage, but it is tied to the public behavior above. When URI schemes become part of the user contract, tests around them are not just housekeeping.

Upgrade notes

The release notes do not mark v1.0.61 as a breaking release and the GitHub metadata says prerelease: false. Still, Vitess and PlanetScale users should review source URI values before rolling it into scheduled jobs.

The reason is simple: this release changes how those backends are named in connector setup. If a pipeline still uses a generic MySQL style source where a dedicated Vitess or PlanetScale scheme is now expected, that is the first place to check.

Also review any alert rules or runbooks that match exact error text from MySQL sources. The fail-fast message cleanup can change what operators see in logs. That is a good cleanup, but brittle log matching has a way of turning good cleanup into a noisy morning.

For everyone else, this looks like a narrow connector release. If you do not use Vitess, PlanetScale, or MySQL based sources, the release notes do not point to a direct user visible change.

Where to get it

• Release page: ingestr v1.0.61
• Project repo: bruin-data/ingestr
• Tag: v1.0.61