Insert workload waits come down to two names, WRITELOG and PAGELATCH_EX. This post runs 16 connections that insert one row at a time and reads their waits with a tested script. It explains what each wait means and where to look for a fix.
sqlcmd vs SSMS: How to Time the Same Loop Fairly
A sqlcmd vs SSMS timing is fair only when both tools run the same file. The clock must sit inside the script. This post times 10,000 inserts three ways with a log table, shows what changed and what didn’t, and lists the settings to match before you compare.
oStress vs SSMS: What Each Tool Is Built to Measure
oStress vs SSMS is not a race, because the tools measure different things. This post times 10,000 inserts in three ways with a clock inside the script, then shows how eight sessions finish the same work faster. It explains what a one-session loop really measures.
oStress Load Test: Run Many Sessions Against SQL Server
An oStress load test runs one query from many connections at once, so you can watch SQL Server under a crowd. This post explains the main switches and runs a small first test with five connections. It also builds the same test with sqlcmd workers, for machines without oStress.
Query Store Status for Every Database in SQL Server
Query Store status for every database takes two queries, and the first is not enough. This post builds four demo databases, shows why the is_query_store_on flag hides frozen and read-only states, and runs one script that reads the real state of each database. It also shows what Query Store costs in memory.
Forcing a Plan in Query Store and Checking That It Held
Use forcing a plan in Query Store with verified IDs, then audit force failures and runtime results every week so the fix stays useful.
Duplicate Statistics: Auto-Created Stats That Shadow Index Statistics
Find duplicate statistics on index-leading columns, compare freshness and sampling, and generate cautious DROP STATISTICS candidates.







