While we're testing out different ways our index structure may be improved, we will want the ability to dry-run on real data.
We should:
Run through the steps of generating a test index which contains some random sample of our real data, using the _reindex api- either via kibana, or directly via curl
Document that process here
Produce a script or other utility for replicating that process with various index mapping choices.
While we're testing out different ways our index structure may be improved, we will want the ability to dry-run on real data. We should: