The following guide assumes that you are familiar with Fluent Bit, if that is not the case we suggest you review the official manual first:
Basic understanding of Structured Query Language (SQL)
A Stream represents an unique flow of data being ingested by an Input plugin. By default Streams get a name using the plugin name plus an internal numerical identification, e.g: tail.0 . Stream name can be changed setting the alias property.
Stream Processor configuration have the notion of Tasks that represents an execution unit, for short: SQL queries are configured in a Task.
When Stream Processor runs a SQL query, results are generated. These results can be re-ingested back into the main Fluent Bit pipeline or simply redirected to the standard output interfaces for debugging purposes.
Fluent Bit group records and associate a Tag to them. Tags are used to define routing rules or in the case of the stream processor to attach to specific Tag that matches a pattern.
Matching rule that can use a wildcard to match specific records associated to a Tag.