sp_updatestats and disabled indexes: the procedure updates statistics on disabled nonclustered indexes and skips a table whose clustered index is disabled. This post proves both, times the cost of disabled indexes against dropped ones on a million rows, and generates a targeted update that skips them.
JOIN Elimination in SQL Server: When a Table Drops Out
JOIN elimination removes a joined table from the plan when the query does not need it. This post proves it with STATISTICS IO, then tests nine variations to show when it applies and when it does not, including a foreign key that has lost its trust.
sp_autostats: See and Undo Automatic Statistics Updates
The sp_autostats procedure shows and changes the automatic statistics update setting of one table, one index or one statistic. This post turns the setting off for a table, flips it back for one index, undoes it, and shows that the database level setting and the statistics themselves are untouched.
SQL SERVER – What is Logical Read?
A logical read requests a database page through the buffer cache. A customer asked about that measure after my STATISTICS IO video.
SQL SERVER – 3 Different Ways to Set MAXDOP
There are 3 different MAXDOP scopes: instance, database and statement. Readers of my 16-versus-1 demonstration asked where to configure it.
SQL SERVER – PREEMPTIVE Waits and Scheduler Context
Preemptive waits can measure work outside SQL Server cooperative scheduling. A customer asked about their increase during a health check.
Multiple OPTIMIZE FOR Hints in One Query: Value or UNKNOWN
Multiple OPTIMIZE FOR hints in one query are written as one list: each parameter gets its own value or UNKNOWN. This post compiles the same query six ways, reads the compile values and estimated rows from the plan cache, and shows which combinations SQL Server accepts.







