We currently have a scripted input into Splunk that is a CSV and we are doing field extractions via regex. This is not super flexible as we want to add fields to the input, but if we change the regex it will work for the new data, but not the historical data. So, instead we want to change our input to be key/value pairs (key="value"). CSV is lean, but inflexible. The key value pairs are more robust, but obviously not very lean. Honestly, we don't care about disk space ... space is cheap. The question is will this affect our daily index license if the keys have the same names as the previous field extractions. Basically, is it file size, or the amount of data that is actually indexed (are these the same thing) that goes toward the daily license? Thanks.
The index volume license is just that. It is the volume of data that is indexed, and it should be roughly equal to the uncompressed size of the log files. If changing to key value pairs increases the size of the log files, then it will also increase the volume indexed.
The index volume license is just that. It is the volume of data that is indexed, and it should be roughly equal to the uncompressed size of the log files. If changing to key value pairs increases the size of the log files, then it will also increase the volume indexed.
In my opinion it should affect the daily index volume as header for each column will now be repeating for each row/event. Daily license goes by data actually indexed.