Here’s a story that teaches a real-world lesson from those internals. The Case of the Midnight Slowdown
UPDATE STATISTICS Orders; The plan switched to an index seek. The ETL dropped to 12 minutes. Good, but not great. Why not 8 minutes? Alex dug deeper. During the ETL, he monitored:
SELECT last_user_seek, last_user_scan, modifications FROM sys.dm_db_index_usage_stats WHERE database_id = DB_ID('SalesDB') AND object_id = OBJECT_ID('Orders'); The result: last_user_seek was yesterday. modifications was over 50,000. Guru Guide To Sql Server Architecture And Internals.pdf
That open transaction was preventing the transaction log from truncating. The log had grown to 200 GB. The ETL’s large update inside FactSales_Load had to wait for log space, causing log autogrowth events (zero-initialization → slow).
I can’t directly open or read the contents of a specific PDF file like Guru Guide To SQL Server Architecture And Internals.pdf . However, I can give you a based on the typical themes found in that book—focusing on SQL Server’s core architecture (query processor, storage engine, buffer pool, transaction log, and locking). Here’s a story that teaches a real-world lesson
SELECT * FROM sys.dm_os_buffer_descriptors WHERE database_id = DB_ID('SalesDB'); He saw that 40 GB of the buffer pool was filled with old data from a morning report. The ETL’s needed pages (the clustered index of Orders ) were being paged in from disk— couldn’t save it because the scan had already caused random I/O earlier.
Index stats were stale. The query optimizer thought the scan was cheaper because it didn’t know the table had grown massively since the last stats update. Good, but not great
He ran:
The buffer pool is a shared resource. Morning report’s KEEP hints or large scans polluted the cache.
He looked at sys.dm_tran_database_transactions during the ETL. One transaction had an old database_transaction_begin_time from 3 hours ago—an open transaction from a developer’s BEGIN TRAN in SSMS that was never committed or rolled back.