This message was deleted.

Yes. One or more segment files are created per time interval in the partitioned by clause. The __time value determine in which interval the row is stored. __time values can be as precise as you want.

A month will (likely) take up multiple segments, but you won't find two months' data in one segment, if that makes sense.

@Sergio Ferragut Datasource segment gets created during the data ingestion process like loading from s3 files or SQL query execution like REPLACE or INSERT statement also or both ways?

All ingestion methods create segments grouped into time intervals called "segment granularity". There are a few ways to ingest data. For batch index_parallel or SQL based ingestion (REPLACE/INSERT), for streaming index_kafka or index_kinesis. All forms of ingestion produce segment files, all segment files belong to a particular time interval. For native batch and streaming jobs the time intervals are defined with segmentGranularity, for SQL based ingestion it is through the PARTITIONED BY clause.

Thank you