I Get Column names using a schema-qualified table and a defined column order. The four methods use corrected filters.
Normalisation in Plain Words, and When to Stop
Understand normalisation through one order example, then decide when a reporting copy is worth the cost of keeping duplicated data consistent.
SQL SERVER – Fix: There is already an object named ‘MyFirstTable’ in the database
Recently I had a very interesting scenario with a client when they faced an error about already object named that exist in the database.
AT TIME ZONE Twice: Convert UTC to Local Time in SQL Server
AT TIME ZONE twice converts UTC to local time in SQL Server. The first use labels a plain value as UTC and the second converts it. A demo shows the common one step mistake, stores in three cities, and the missing and repeated hours of daylight saving.
Escape Text for JSON: Build Valid JSON With STRING_ESCAPE
To escape text for JSON, pass each value through STRING_ESCAPE, so JSON built by hand stays valid. A demo shows which characters it escapes and which it leaves alone. It also shows when JSON_MODIFY and FOR JSON make it unnecessary.
Forwarded Records and Performance – SQL in Sixty Seconds #155
Forwarded records make SQL Server read extra pages when it scans a heap. In this SQL in Sixty Seconds video, I explain forwarded records and performance.
Which Tables Are Huge: Rows Versus Size in SQL Server
Rows versus size decides which tables are huge in SQL Server. One query reads sys.dm_db_partition_stats and ranks every table twice, by rows and by size. A demo shows 300 rows that outweigh 300,000 rows, and why the older join misleads.