https://logstag.com/blog/ Notes on keeping the data layer reliable: for each database, and on how Logstag is built. en https://logstag.com/blog/database-reliability-is-not-uptime/ https://logstag.com/blog/database-reliability-is-not-uptime/ Sun, 04 Oct 2026 09:00:00 +0000 A database can pass every health check and still fail its users. Five kinds of signals that show whether your data layer is reliable, plus a checklist. https://logstag.com/blog/one-finding-167-times/ https://logstag.com/blog/one-finding-167-times/ Sun, 04 Oct 2026 09:00:00 +0000 A detector that never accepts a new normal ends up reporting itself, not the database. How we taught Insights to tell news from a new normal, and to ignore changes too small to matter. https://logstag.com/blog/oracle-ash/ https://logstag.com/blog/oracle-ash/ Sun, 04 Oct 2026 09:00:00 +0000 What Active Session History is, why you count samples instead of summing wait times, three ASH queries, and what you can see without the Diagnostics Pack. https://logstag.com/blog/sql-server-waits/ https://logstag.com/blog/sql-server-waits/ Sun, 04 Oct 2026 09:00:00 +0000 Read sys.dm_os_wait_stats as deltas, filter benign waits, learn what LCK_M, PAGEIOLATCH, WRITELOG and CXPACKET point to, and find the head blocker.