Standardized production runbooks for PostgreSQL, MySQL, SQL Server, Oracle, and MongoDB: Symptom, Impact, Diagnosis, Response, Verification, and Rollback.
Walks through resolving rejected new connections when the limit is reached. Finds the responsible app, frees slots safely and covers the permanent fix.
Steps to save a database from going read-only when XID age reaches a critical level. Clears blockers and runs VACUUM FREEZE in the right order.
Finds the session at the head of a lock chain and resolves it safely. Covers the DDL queue trap and prevention methods.
Separates slow queries, lock waits, disk reads and temp-file spills to guide the right response.
Separately measures WAL generation, transfer, flush and replay to identify network, disk or query causes of lag.
Separates WAL growth due to slots, archiving, replicas and write volume and gives safe recovery steps without deleting WAL files manually.
Diagnoses dead tuple growth, evaluates autovacuum triggers and uses safe per-table tuning to prevent recurring bloat.
Diagnoses open transactions by age, owner, locks and xmin, and safely resolves sessions blocking vacuum.
Combines deadlock counters and logs to identify conflicting transactions and objects, then prevents recurrence through lock ordering and index design.
Diagnoses requested checkpoints and write bursts with version-appropriate statistics, then balances WAL capacity and checkpoint pacing.
