Hi,

I have built a pipeline performing reads using ElasticsearchIO  from a
large Elasticsearch index version 6.5.3. While it seems to run just fine
using Flink runners there is absolutely no feedback in terms of progress.
Am I correct in saying that this connector has no progress reporting
implemented ?

Could someone also please confirm that this connector has to read the
entire query response before the data gets passed to the next PTransform ?
Is there a way to process the data un chunks to avoid the lack of
visibility into the progress of the query ?

Because there is a lack of any sort of metrics it also seems to imply that
the runners cannot scale automatically because they have no info. I do see
however in the source code that query is split into bundles and as of ES
version 5.0 a fair bit of functionality for parallel querying is provided.

Kind Thanks,
Serge

Reply via email to