Hi all,
Is their a formula to calculate % Utilization for Indexes? Where the formula is using VSTs
Right now I'm using a script that runs the DBANALYS command with the -csoutput option and parsing through the index file to retrieve the % Utilization for each Index. I think that using a program that calculates this would be more efficient.
From looking at the manual and other posts it seems like the only ways to view % Utilization is with the DBANALYS or IDXANALYS commands. But maybe I'm looking in the wrong direction.
Your current approach sounds reasonable. Is there a problem with it that makes you want to change?
I think you'd be hard-pressed to find a better way to do this from the ABL. I don't know of any VST info that will give you insight into b-tree content.
It's a good idea to be running regular dbanalys reports anyway; say, monthly. I wouldn't expect your index utilization to change so quickly or dramatically that parsing that monthly report wouldn't be good enough.
I want to change my current approach in case there is ever an upgrade or update that would break my current script. Say if Progress ever changes the formatting/layout of the csoutput. VSTs just seem more consistent or secure.
With -csoutput, you can add -verbose to include the column names before in the first record. So if a new column is inserted in one of the .txt files, you'll still know what's what. That should make parsing reliable enough.