Earlier I wrote a blog post about how Data Compression for Performance can help improve the efficiency of SQL Server.
CPU Running 100% – SQL in Sixty Seconds #185
I got lots of responses to this blog and people wanted me to create a video for this topic. Here, I have a video for CPU Running 100%.
Aggregate Pushdown: Letting the Columnstore Scan Do the Summing
Verify aggregate pushdown in columnstore scans, inspect locally aggregated rows, and compare numeric aggregates with expressions.
Index a Computed Column in SQL Server
To index a computed column, define the formula once on the table and create a normal index on it. This post loads 200,000 rows, shows the scan, then the seek after the index, and tests the rules that decide whether SQL Server accepts the column.
Reading the Connectivity Ring Buffer for Dropped Connections
The error log says nothing, but the connectivity ring buffer may remember. Here is how to read it and what it cannot tell you.
Status Changes Only: Collapsing Repeated Rows With LAG
Hundreds of repeated readings, three real changes. Here is how LAG finds them, and why a plain not-equal test quietly drops rows.
Latch Waits at the Top: Reading sys.dm_os_latch_stats
Sample SQL Server latch waits by class, then connect ACCESS_METHODS_DATASET_PARENT, FGCB_ADD_REMOVE, and LOG_MANAGER to the next check.