Hi Israel, I don't see BigTableIO APIs changing at this point. So I won't call it truly experimental. However we have another BigTable connector that uses the HBase API available here. https://github.com/googleapis/java-bigtable-hbase/tree/master/bigtable-dataflow-parent/bigtable-hbase-beam
This one was developed and is supported by the Cloud BigTable team. So I recommend using that one unless you have a strong reason to use the one in Beam. Thanks, Cham On Mon, Jan 20, 2020 at 3:07 PM Israel Herraiz <[email protected]> wrote: > Hi, > > I have been working lately quite a lot with Dataflow and Bigtable, using > Beam's BigtableIO. > > The documentation ( > https://beam.apache.org/releases/javadoc/2.17.0/org/apache/beam/sdk/io/gcp/bigtable/BigtableIO.html) > shows that BigtableIO is annotated as experimental. > > The same docs also says that: *this connector for Cloud Bigtable is > considered experimental and may break or receive backwards-incompatible > changes in future versions of the Apache Beam SDK. Cloud Bigtable is in > Beta, and thus it may introduce breaking changes in future revisions of its > service or APIs.* > > However, Bigtable went out of beta in 2016 ( > https://cloud.google.com/bigtable/docs/release-notes#June_29_2016) and > many of the client libraries (all?) are in GA too (Java went out of beta on > September -> > https://cloud.google.com/bigtable/docs/release-notes#September_25_2019). > > So I am wondering if BigtableIO is truly experimental, and whether I > should submit a pull request updating the comment about Cloud Bigtable > being in beta. > > Kind regards, > Israel >
