The current Cloud SQL Python connector is great. But in it's current form, when used via Apache Beam/Google Cloud Dataflow, it is not recognized as an official Apache Beam framework construct. This can lead to unbalanced workers in Dataflow when dealing with any kind of volume in a pipeline.
I suggest creating an official Beam I/O Connector so that Cloud SQL can have optimized usage with Beam/Dataflow.
Sample code
No response
Alternatives Considered
There is current non-native Cloud SQL support, but it requires a custom container. This can cause slowdowns and longer approval processes at regulated customers. An official native I/O connector would be ideal.
Feature Description
The current Cloud SQL Python connector is great. But in it's current form, when used via Apache Beam/Google Cloud Dataflow, it is not recognized as an official Apache Beam framework construct. This can lead to unbalanced workers in Dataflow when dealing with any kind of volume in a pipeline.
I suggest creating an official Beam I/O Connector so that Cloud SQL can have optimized usage with Beam/Dataflow.
Sample code
No response
Alternatives Considered
There is current non-native Cloud SQL support, but it requires a custom container. This can cause slowdowns and longer approval processes at regulated customers. An official native I/O connector would be ideal.
Additional Details
No response