How to quickly delete a very large table in MySQL
MySQL Quick deletion of very large tables
If you delete a large table directly in MySQL, it may cause MySQL to hang, which may have a negative impact on your business. The prerequisite for deleting a very large table is that the table is an independent table space, so that the deletion is effective.
Table creates a hard link
# du -sh pay_bills.ibd 175G pay_bills.ibd # 创建硬链接 # ln pay_bills.ibd pay_bills.ibd_hdlk
Perform table deletion
In Linux, each stored file will have an Inode Index pointing to the file, and multiple file names can be passed The same Inode Index points to the same storage file.
If the Inode Index referenced by the file name is also referenced by other file names, only the reference between the file name and the Inode Index will be deleted
If the Inode Index referenced by the file name is If it is not referenced by other file names, delete the reference between the file name and the Inode Index and delete the storage file pointed to by the Inode Index.
In fact, only a file reference to pay_bills.ibd is deleted. The reference to the physical file of pay_bills.ibd_hdlk still exists, so no OS-level deletion operation will be performed. The IO fluctuation is not large, which reduces the need for MySQL impact.
mysql> drop table pay_bills; Query OK, 0 rows affected (3.24 sec)
Execute file deletion
Install the truncate tool
# yum install coreutils -y
Execute the deletion script
#!/bin/bash TRUNCATE=/usr/bin/truncate # 从175G开始每次删除2G,最后如果脚本truncate后还剩下部分文件,使用rm删除 for i in `seq 175 -2 1`; do $TRUNCATE -s ${i}G pay_bills.ibd_hdlk sleep 1 done rm -f pay_bills.ibd_hdlk
MySQL quickly clears large table data
First time for the project When going online, the amount of data created for performance testing is huge, and some of it is unusable, but some of it must be retained. In many cases, it is necessary to systematically clean the data or filter out the useful data before inserting it into the table. middle! Keep the table structure or rebuild the table (export the table structure in the database), and re-execute the SQL statement.
MySQL deletes or clears table data
Five ways to clear table data
1. Truncate– deletes all data and retains the table structure. It cannot Undo and restore
2. delete– is extremely slow to delete row by row and is not suitable for deleting large amounts of data
3. drop– deletes the table, and the data and table structure are deleted together
4. Export the table structure and execute it again;
5. Export all table structures of the database and execute it again to clear all tables (consistent with 4)
Clear table data syntax
truncate table table name; (only the table structure is retained, very fast)
delete from table name;
delete from table name where column name="value ";
DROP `t_product_events` IF EXIST; CREATE TABLE `t_product_events` ( `id` bigint(20) UNSIGNED NOT NULL AUTO_INCREMENT COMMENT '事件ID', `level` int(11) NULL DEFAULT NULL, `product_id` bigint(20) NOT NULL COMMENT '产品类型ID', `identifier` varchar(64) CHARACTER SET utf8 COLLATE utf8_general_ci NULL DEFAULT NULL, `name` varchar(64) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NOT NULL COMMENT '事件名称', `description` varchar(256) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NULL DEFAULT NULL COMMENT '事件描述', `type` int(11) NOT NULL COMMENT '事件类型,0:info(信息)、1:alert(告警)、2:error(故障)', `ref_id` int(11) UNSIGNED NULL DEFAULT 0 COMMENT '引入模板时有意义', `original_required` tinyint(2) NOT NULL, `update_required` tinyint(2) NOT NULL DEFAULT 0 COMMENT '是否是标准功能的必选事件,0:可选,1:必选', `custom` tinyint(2) NOT NULL COMMENT '0:模板导入,1:自定义', `method` varchar(128) CHARACTER SET utf8mb4 COLLATE utf8mb4_general_ci NOT NULL COMMENT '事件对应的方法名称(根据identifier生成)', `create_time` datetime(0) NOT NULL DEFAULT CURRENT_TIMESTAMP, `update_time` datetime(0) NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP(0), `ref` bigint(20) NULL DEFAULT NULL, `related` int(11) NULL DEFAULT 0 COMMENT '被预发布或者已发布关联个数', `copyright` tinyint(2) NULL DEFAULT 0 COMMENT '是否发布过, 1 发布过 0未发布', `prerelease` tinyint(2) NULL DEFAULT 0 COMMENT '是否预发布过, 1 预发布过 0未预发布', PRIMARY KEY (`id`) USING BTREE, INDEX `idx_product_id`(`product_id`) USING BTREE COMMENT '查询优化' ) ENGINE = InnoDB AUTO_INCREMENT = 497560 CHARACTER SET = utf8mb4 COLLATE = utf8mb4_general_ci COMMENT = '产品类型事件表' ROW_FORMAT = Compact;
The above is the detailed content of How to quickly delete a very large table in MySQL. For more information, please follow other related articles on the PHP Chinese website!

Hot AI Tools

Undress AI Tool
Undress images for free

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

Notepad++7.3.1
Easy-to-use and free code editor

SublimeText3 Chinese version
Chinese version, very easy to use

Zend Studio 13.0.1
Powerful PHP integrated development environment

Dreamweaver CS6
Visual web development tools

SublimeText3 Mac version
God-level code editing software (SublimeText3)

To add a primary key to an existing table, use the ALTERTABLE statement with the ADDPRIMARYKEY clause. 1. Ensure that the target column has no NULL value, no duplication and is defined as NOTNULL; 2. The single-column primary key syntax is ALTERTABLE table name ADDPRIMARYKEY (column name); 3. The multi-column combination primary key syntax is ALTERTABLE table name ADDPRIMARYKEY (column 1, column 2); 4. If the column allows NULL, you must first execute MODIFY to set NOTNULL; 5. Each table can only have one primary key, and the old primary key must be deleted before adding; 6. If you need to increase it yourself, you can use MODIFY to set AUTO_INCREMENT. Ensure data before operation

B-TreeindexesarebestformostPHPapplications,astheysupportequalityandrangequeries,sorting,andareidealforcolumnsusedinWHERE,JOIN,orORDERBYclauses;2.Full-Textindexesshouldbeusedfornaturallanguageorbooleansearchesontextfieldslikearticlesorproductdescripti

Using mysqldump is the most common and effective way to back up MySQL databases. It can generate SQL scripts containing table structure and data. 1. The basic syntax is: mysqldump-u[user name]-p[database name]>backup_file.sql. After execution, enter the password to generate a backup file. 2. Back up multiple databases with --databases option: mysqldump-uroot-p--databasesdb1db2>multiple_dbs_backup.sql. 3. Back up all databases with --all-databases: mysqldump-uroot-p

You can customize the separator by using the SEPARATOR keyword in the GROUP_CONCAT() function; 1. Use SEPARATOR to specify a custom separator, such as SEPARATOR'; 'The separator can be changed to a semicolon and plus space; 2. Common examples include using the pipe character '|', space'', line break character '\n' or custom string '->' as the separator; 3. Note that the separator must be a string literal or expression, and the result length is limited by the group_concat_max_len variable, which can be adjusted by SETSESSIONgroup_concat_max_len=10000; 4. SEPARATOR is optional

UNIONremovesduplicateswhileUNIONALLkeepsallrowsincludingduplicates;1.UNIONperformsdeduplicationbysortingandcomparingrows,returningonlyuniqueresults,whichmakesitsloweronlargedatasets;2.UNIONALLincludeseveryrowfromeachquerywithoutcheckingforduplicates,

The table can be locked manually using LOCKTABLES. The READ lock allows multiple sessions to read but cannot be written. The WRITE lock provides exclusive read and write permissions for the current session and other sessions cannot read and write. 2. The lock is only for the current connection. Execution of STARTTRANSACTION and other commands will implicitly release the lock. After locking, it can only access the locked table; 3. Only use it in specific scenarios such as MyISAM table maintenance and data backup. InnoDB should give priority to using transaction and row-level locks such as SELECT...FORUPDATE to avoid performance problems; 4. After the operation is completed, UNLOCKTABLES must be explicitly released, otherwise resource blockage may occur.

To select data from MySQL table, you should use SELECT statement, 1. Use SELECTcolumn1, column2FROMtable_name to obtain the specified column, or use SELECT* to obtain all columns; 2. Use WHERE clause to filter rows, such as SELECTname, ageFROMusersWHEREage>25; 3. Use ORDERBY to sort the results, such as ORDERBYageDESC, representing descending order of age; 4. Use LIMIT to limit the number of rows, such as LIMIT5 to return the first 5 rows, or use LIMIT10OFFSET20 to implement paging; 5. Use AND, OR and parentheses to combine

TheINoperatorinMySQLchecksifavaluematchesanyinaspecifiedlist,simplifyingmultipleORconditions;itworkswithliterals,strings,dates,andsubqueries,improvesqueryreadability,performswellonindexedcolumns,supportsNOTIN(withcautionforNULLs),andcanbecombinedwith
