This guide explains how to regularly purge oversized log tables using time‑based DELETE statements, interprets related cloud RDS and ECS alerts, and offers troubleshooting steps and batch‑deletion techniques to address MySQL read‑replica synchronization delays caused by large transactions.
This article explains how MySQL's offline_mode variable and the pt‑kill tool can be used by DBAs to place a server in maintenance mode, instantly dropping non‑privileged client connections, preventing new ones, and providing a safe way to perform maintenance or backups.
This article analyzes a large MySQL table with frequent SLA alerts, demonstrates how to identify slow queries using pt‑query‑digest, proposes index redesign and online DDL versus pt‑osc methods, and shows testing, implementation, and batch‑delete techniques to dramatically reduce query time and maintenance overhead.
This article details a comprehensive analysis of a large MySQL table's performance issues, including slow query diagnostics with pt‑query‑digest, index evaluation, backup and restore procedures, online DDL versus pt‑osc, and practical strategies for index redesign and batch deletions to reduce latency and maintenance overhead.
This article explains the cause of PostgreSQL "database is not accepting commands to avoid wraparound data loss" errors, details the XID allocation and wraparound mechanisms, and provides manual and automated vacuum‑freeze solutions with concrete SQL commands and scripts.
The article explains why MySQL’s DELETE command does not shrink the physical table file, describes InnoDB’s storage mechanics, and shows how to reclaim space using OPTIMIZE TABLE, ALTER TABLE, and Online DDL options.
The article explains why deleting rows in MySQL does not shrink the physical table file, describes the underlying InnoDB storage mechanisms, and provides practical methods such as OPTIMIZE TABLE, ALTER TABLE, and Online DDL to reclaim space and reorganize the table.
This article walks through diagnosing MySQL disk‑usage spikes caused by data fragmentation, explains why common cleanup commands may fail, and provides a detailed, repeatable procedure—including stopping the service, removing InnoDB files, and restoring the database—to finally free up the occupied space.
This article explains how to detect table‑space fragmentation in MySQL, demonstrates the performance impact of deleted rows, and provides step‑by‑step commands—including ALTER TABLE FORCE, OPTIMIZE TABLE, and mysqlcheck—to reclaim space and improve query speed.
This article, the fourth in a MySQL fundamentals series, explains how to diagnose and optimize slow SQL statements using SHOW STATUS, the slow‑query log, SHOW PROCESSLIST, and EXPLAIN, then details index types, creation, usage rules, and how to analyze, check, and optimize tables with MySQL commands.
The article explains why deleting rows in MySQL does not shrink the .ibd file because InnoDB only marks pages as reusable, and shows how to reclaim space using OPTIMIZE TABLE, ALTER TABLE, and Online DDL options.
This article explains MongoDB TTL indexes, covering their basic concept as single‑field auto‑deletion indexes, how the background process works, alternative creation methods using an expireAt field, practical limitations, and recommendations for designing efficient data expiration strategies.
This guide walks you through the essential daily maintenance tasks for GaussDB T clusters, covering ETCD startup, cluster health checks, host resource monitoring, tablespace usage, abnormal wait events, log inspection, and common error troubleshooting with concrete commands and SQL examples.
This guide explains why MySQL tables become fragmented after deletions, how to check the Data_free value to identify fragmented tables, and which commands to run for MyISAM and InnoDB tables to reclaim space and improve query performance.
Dual‑master MySQL replication lets two servers act as each other's master and slave, preventing circular replication through unique server‑ids, reducing maintenance overhead, enabling automatic failover, and allowing only one node to handle writes to avoid data conflicts, thereby minimizing downtime.
