Splunk extracts metadata fields such as host, source, and sourcetype when it indexes events. This process helps in identifying and organizing the data, making it easier to search and analyze later. While indexing, Splunk does not limit extraction to just the top 10 fields, nor does it perform parsing, merging, and typing processes on universal forwarders. Also, creating report acceleration summaries is a different function and not a part of the indexing process.