Using ingestion actions, one can write a copy of events to an S3 bucket prior to indexing. Can one search these S3 buckets with Splunk even though they were not ingested (it'd be slow, but could be useful for historical searches)?
