I have a transform that uses a query to initially filter some of my data by a 'state' field (e.g values of state_1, state_2 etc.) before creating the aggregation, the state can change e.g from state_3 to state_2 but this data is not in the transform.
If I delete the transform and start it again it will include the updated data, is this how a transform is supposed to work?
Each time the continuous transform is refreshed is the query run again on all data or is the query only run once when the transform is started to create an initial subset?
It's the former. For continuous transforms, it's part of the process where the transform is checking for changes to source indices. There's some information about these checkpoints in the documentation:
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.