Jeff Moden wrote:
For my databases, I've adopted the "If it moves, shoot it" philosophy for statistics updpates. In other words, if the RowModCntr is not zero and the stats haven't been updated in the last 23 hours, I update the individual stats with a FULL SCAN. I do that every night on my most active databases and once a week on those that are less active.
That's a nice luxury to have. I can't afford the hit to the buffer cache to fully scan all my active, large tables every night (billions and billions of rows). That would definitely blow out the existing cache and massively degrade the PLE. Which may or may not be critical, depending on your box, but usually will be, since most shops are vastly more I/O bound than CPU bound.
Edit: And, with further thought, there's no chance such a process could finish fully scanning all my updated tables within one night. We have hundreds of clients, many of whom individually have billions and billions of rows.
SQL DBA,SQL Server MVP(07, 08, 09) Prosecutor James Blackburn, in closing argument in the Fatal Vision murders trial: "If in the future, you should cry a tear, cry one for them [the murder victims]. If in the future, you should say a prayer, say one for them. And if in the future, you should light a candle, light one for them."