On Feb 1, 2016, at 6:58 PM, David G. Johnston wrote:
Yeah, i didn't think there would be support. The stats collector doesn't have the info that I want... it's focused on how the data is used. I'm more interested in what the data is. Basically I want to compare the distribution of index "keys". In the case of a substring index, comparing the distribution at 3,4,5,6,7,8 characters. based on that, i can run some server tests on different lengths, and the stats collector comes into play. i'm taking a blind stab on some index contents, and want to be a bit more educated. anyways, I eventually realized that I may be better just pulling the columns and running some offline analytics (it could be done in sql, but the results would be nicer as a graph). |
