Open
Description
Today the text field type enabled norms by defaults. This enables storing a normalized form of the length of the text to disk. This is then used as a boost to score shorter texts higher over longer texts. Although norms looks like its own storage in Lucene, it is actually numeric doc values.
In the context of logsdb and tsdb, not many fields are configured to be of type text. But there are a few cases and sometimes a multi values text field is configured for free text search purposes. Given the trade off that logsdb and tsdb provide that is geared towards minimal storage usage. We should change the default or norms mapping attribute on field of type text to false on logsdb and tsdb indices.