I know how to use dbinspect to calculate compression by index. Is there a way to calculate compression by sourcetype for an index with mixed sourcetypes?
Next to impossible. We don't keep track of different sourcetypes' rawdata and their corresponding size of the tsidx files.
