Database Wait Event Diagnostician

Wait event analysis is the most direct path to understanding why a database is slow. When a database session can't proceed — because it's waiting for a lock, for I/O to complete, for CPU time, or for memory — it records that wait. Aggregating and interpreting these waits tells you precisely where the system is spending its time and what is preventing faster execution. This AI assistant is built for DBAs who want to use wait event data as their primary diagnostic tool for performance analysis.

The assistant covers wait event analysis across the major database platforms in depth. For Oracle, it addresses the Automatic Workload Repository (AWR) and Active Session History (ASH) in detail — how to read a Top Timed Events section in an AWR report, what each common wait class means (User I/O, System I/O, Concurrency, Cluster, Application), and how to use ASH data to drill down into which SQL statements and sessions are responsible for the waits. It explains the difference between wait events that indicate a root cause versus those that are symptoms of another problem.

For SQL Server, the assistant covers sys.dm_os_wait_stats and sys.dm_exec_session_wait_stats, explaining the most diagnostically significant wait types — PAGEIOLATCH (storage I/O), LCK (lock contention), CXPACKET/CXCONSUMER (parallelism), WRITELOG (log flush), SOS_SCHEDULER_YIELD (CPU pressure), and ASYNC_NETWORK_IO (network or application issues). For PostgreSQL, it addresses pg_stat_activity wait events, pg_stat_bgwriter, and lock wait visibility.

The assistant helps you build a systematic wait analysis methodology — establishing a baseline, capturing wait data during a performance problem, comparing the two, and using the delta to isolate the bottleneck. It explains how to avoid common misinterpretations, such as confusing high-volume benign waits with performance-impacting waits. This is an essential tool for DBAs conducting performance investigations, platform engineers diagnosing production incidents, and consultants preparing database health assessments.