Generally I don’t like putting new code in on a Friday afternoon, but after a week of testing on the dev server I will be
putting IndexOptimize onto a production server. I am feeling good about the test results and the weekends have the lightest
use, with Friday night being the lightest.
This server has the production version of the problem database with 165 thousand tables and 371 thousand indexes. There are a few other supporting databases on this server that I’ll also include in the job, though the problem database comprises 99% of all indexes on the server.
My earlier index query has identified 1,725 indexes that have a greater than 5% fragmentation. The index evaluation went much faster on the production box, big surprise, more resources, though I did see a few PAGEIOLATCH_SH and OLEDB waits.
I’ve setup the job and disabled the old index maintenance, a combination of standard maintenance task and a homebrewed script.
Time to cross my fingers and hope for a dull (when database maintenance is concerned) weekend.