One of our customers had problems maintaining some monster databases (VLDB). These have about 2 to 4 TB of data and more than 90,000 tables. Building a good maintenance solution for such databases isn’t easy. For most of my clients I use the Ola Hallengren solution, which has proved to be excellent before but it just wasn’t suitable here. The first thing on my list was getting the databases checked. But the normal DBCC CHECKDB command went on for days until its snapshot clogged up my server.