In my post Row and Page Count Influence on Estimated CPU and I/O Cost I demonstrated how I/O cost of a clustered index scan had sensitivity to page counts but not row counts.  For this post I’ll lay out a direct connection between data page counts and the query execution plan that gets generated. To [...]

Last November I blogged about how index usage stats don’t get updated when the associated index statistics (but not index) are used. This post will describe another scenario that you should be aware of (the topic came up today in class while Kimberly was teaching – as we were trying to recall tricks to clearing [...]

I was thinking back to a conversation I had some time ago regarding cases when sys.dm_db_index_usage_stats does not get updated after an index is used for just the statistics associated with the indexed columns.  What makes this profession so interesting is that almost every tool we use has a "yes, but…" associated with it.  What [...]

Today there isn’t a fine-tuned way to clear the accumulated statistics from the “missing index” set of DMVs.  Certainly restarting SQL Server will do this, but you don’t have something like DBCC SQLPERF (‘sys.dm_os_wait_stats’, CLEAR) to reset statistics for dm_db_missing_index_columns, dm_db_missing_index_groups, dm_db_missing_index_group_stats and dm_db_missing_index_details. While we don’t have this direct option, there is an interesting [...]