The CirrusSearchRequestSet schema needs to have a timestamp field for the imported data to be partitioned correctly. The schema needs to be updated in two places:
This is the schema on the analytics end (note that it has a different namespace, only as a temporary workaround on the import side of things)
- On the monolog end (you can use the search namespace here - no problem).
To update on the analytics end, send us an updated schema version and we can update it/submit a patch to refinery-source and we can review and merge.