温馨提示×

温馨提示×

您好,登录后才能下订单哦!

密码登录×
登录注册×
其他方式登录
点击 登录注册 即表示同意《亿速云用户服务条款》

zabbix上怎么对mysql数据库做分区表

发布时间:2021-11-10 11:27:56 来源:亿速云 阅读:253 作者:iii 栏目:MySQL数据库

本篇内容介绍了“zabbix上怎么对mysql数据库做分区表”的有关知识,在实际案例的操作过程中,不少人都会遇到这样的困境,接下来就让小编带领大家学习一下如何处理这些情况吧!希望大家仔细阅读,能够学有所成!

独立表空间设置:

innodb_file_per_table=1

修改表索引:

zabbix3.2以上版本跳过此步骤

mysql> Alter table history_text drop primary key, add index (id), drop index history_text_2, add index history_text_2 (itemid, id);

Query OK, 0 rows affected (0.49 sec)

Records: 0  Duplicates: 0  Warnings: 0

mysql> Alter table history_log drop primary key, add index (id), drop index history_log_2, add index history_log_2 (itemid, id);

Query OK, 0 rows affected (2.71 sec)

Records: 0  Duplicates: 0  Warnings: 0

mysql>

创建存储过程:

增加分区存储过程:

DELIMITER $$CREATE PROCEDURE `partition_create`(SCHEMANAME varchar(64), TABLENAME varchar(64), PARTITIONNAME varchar(64), CLOCK int)BEGIN
        /*           SCHEMANAME = The DB schema in which to make changes           TABLENAME = The table with partitions to potentially delete           PARTITIONNAME = The name of the partition to create        */
        /*           Verify that the partition does not already exist        */
        DECLARE RETROWS INT;
        SELECT COUNT(1) INTO RETROWS
        FROM information_schema.partitions
        WHERE table_schema = SCHEMANAME AND table_name = TABLENAME AND partition_description >= CLOCK;
        IF RETROWS = 0 THEN
                /*                   1. Print a message indicating that a partition was created.                   2. Create the SQL to create the partition.                   3. Execute the SQL from #2.                */
                SELECT CONCAT( "partition_create(", SCHEMANAME, ",", TABLENAME, ",", PARTITIONNAME, ",", CLOCK, ")" ) AS msg;
                SET @sql = CONCAT( 'ALTER TABLE ', SCHEMANAME, '.', TABLENAME, ' ADD PARTITION (PARTITION ', PARTITIONNAME, ' VALUES LESS THAN (', CLOCK, '));' );
                PREPARE STMT FROM @sql;
                EXECUTE STMT;
                DEALLOCATE PREPARE STMT;
        END IF;END$$DELIMITER ;

删除分区存储过程:

DELIMITER $$CREATE PROCEDURE `partition_drop`(SCHEMANAME VARCHAR(64), TABLENAME VARCHAR(64), DELETE_BELOW_PARTITION_DATE BIGINT)BEGIN
        /*           SCHEMANAME = The DB schema in which to make changes           TABLENAME = The table with partitions to potentially delete           DELETE_BELOW_PARTITION_DATE = Delete any partitions with names that are dates older than this one (yyyy-mm-dd)        */
        DECLARE done INT DEFAULT FALSE;
        DECLARE drop_part_name VARCHAR(16);
        /*           Get a list of all the partitions that are older than the date           in DELETE_BELOW_PARTITION_DATE.  All partitions are prefixed with           a "p", so use SUBSTRING TO get rid of that character.        */
        DECLARE myCursor CURSOR FOR
                SELECT partition_name
                FROM information_schema.partitions
                WHERE table_schema = SCHEMANAME AND table_name = TABLENAME AND CAST(SUBSTRING(partition_name FROM 2) AS UNSIGNED) < DELETE_BELOW_PARTITION_DATE;
        DECLARE CONTINUE HANDLER FOR NOT FOUND SET done = TRUE;
        /*           Create the basics for when we need to drop the partition.  Also, create           @drop_partitions to hold a comma-delimited list of all partitions that           should be deleted.        */
        SET @alter_header = CONCAT("ALTER TABLE ", SCHEMANAME, ".", TABLENAME, " DROP PARTITION ");
        SET @drop_partitions = "";
        /*           Start looping through all the partitions that are too old.        */
        OPEN myCursor;
        read_loop: LOOP
                FETCH myCursor INTO drop_part_name;
                IF done THEN
                        LEAVE read_loop;
                END IF;
                SET @drop_partitions = IF(@drop_partitions = "", drop_part_name, CONCAT(@drop_partitions, ",", drop_part_name));
        END LOOP;
        IF @drop_partitions != "" THEN
                /*                   1. Build the SQL to drop all the necessary partitions.                   2. Run the SQL to drop the partitions.                   3. Print out the table partitions that were deleted.                */
                SET @full_sql = CONCAT(@alter_header, @drop_partitions, ";");
                PREPARE STMT FROM @full_sql;
                EXECUTE STMT;
                DEALLOCATE PREPARE STMT;
                SELECT CONCAT(SCHEMANAME, ".", TABLENAME) AS `table`, @drop_partitions AS `partitions_deleted`;
        ELSE
                /*                   No partitions are being deleted, so print out "N/A" (Not applicable) to indicate                   that no changes were made.                */
                SELECT CONCAT(SCHEMANAME, ".", TABLENAME) AS `table`, "N/A" AS `partitions_deleted`;
        END IF;END$$DELIMITER ;

维护分区存储过程:

DELIMITER $$CREATE PROCEDURE `partition_maintenance`(SCHEMA_NAME VARCHAR(32), TABLE_NAME VARCHAR(32), KEEP_DATA_DAYS INT, HOURLY_INTERVAL INT, CREATE_NEXT_INTERVALS INT)BEGIN
        DECLARE OLDER_THAN_PARTITION_DATE VARCHAR(16);
        DECLARE PARTITION_NAME VARCHAR(16);
        DECLARE OLD_PARTITION_NAME VARCHAR(16);
        DECLARE LESS_THAN_TIMESTAMP INT;
        DECLARE CUR_TIME INT;
        CALL partition_verify(SCHEMA_NAME, TABLE_NAME, HOURLY_INTERVAL);
        SET CUR_TIME = UNIX_TIMESTAMP(DATE_FORMAT(NOW(), '%Y-%m-%d 00:00:00'));
        SET @__interval = 1;
        create_loop: LOOP
                IF @__interval > CREATE_NEXT_INTERVALS THEN
                        LEAVE create_loop;
                END IF;
                SET LESS_THAN_TIMESTAMP = CUR_TIME + (HOURLY_INTERVAL * @__interval * 3600);
                SET PARTITION_NAME = FROM_UNIXTIME(CUR_TIME + HOURLY_INTERVAL * (@__interval - 1) * 3600, 'p%Y%m%d%H00');
                IF(PARTITION_NAME != OLD_PARTITION_NAME) THEN
                        CALL partition_create(SCHEMA_NAME, TABLE_NAME, PARTITION_NAME, LESS_THAN_TIMESTAMP);
                END IF;
                SET @__interval=@__interval+1;
                SET OLD_PARTITION_NAME = PARTITION_NAME;
        END LOOP;
        SET OLDER_THAN_PARTITION_DATE=DATE_FORMAT(DATE_SUB(NOW(), INTERVAL KEEP_DATA_DAYS DAY), '%Y%m%d0000');
        CALL partition_drop(SCHEMA_NAME, TABLE_NAME, OLDER_THAN_PARTITION_DATE);END$$DELIMITER ;

检查分区、创建第一个分区存储过程:

DELIMITER $$CREATE PROCEDURE `partition_verify`(SCHEMANAME VARCHAR(64), TABLENAME VARCHAR(64), HOURLYINTERVAL INT(11))BEGIN
        DECLARE PARTITION_NAME VARCHAR(16);
        DECLARE RETROWS INT(11);
        DECLARE FUTURE_TIMESTAMP TIMESTAMP;
        /*         * Check if any partitions exist for the given SCHEMANAME.TABLENAME.         */
        SELECT COUNT(1) INTO RETROWS
        FROM information_schema.partitions
        WHERE table_schema = SCHEMANAME AND table_name = TABLENAME AND partition_name IS NULL;
        /*         * If partitions do not exist, go ahead and partition the table         */
        IF RETROWS = 1 THEN
                /*                 * Take the current date at 00:00:00 and add HOURLYINTERVAL to it.  This is the timestamp below which we will store values.                 * We begin partitioning based on the beginning of a day.  This is because we don't want to generate a random partition                 * that won't necessarily fall in line with the desired partition naming (ie: if the hour interval is 24 hours, we could                 * end up creating a partition now named "p201403270600" when all other partitions will be like "p201403280000").                 */
                SET FUTURE_TIMESTAMP = TIMESTAMPADD(HOUR, HOURLYINTERVAL, CONCAT(CURDATE(), " ", '00:00:00'));
                SET PARTITION_NAME = DATE_FORMAT(CURDATE(), 'p%Y%m%d%H00');
                -- Create the partitioning query
                SET @__PARTITION_SQL = CONCAT("ALTER TABLE ", SCHEMANAME, ".", TABLENAME, " PARTITION BY RANGE(`clock`)");
                SET @__PARTITION_SQL = CONCAT(@__PARTITION_SQL, "(PARTITION ", PARTITION_NAME, " VALUES LESS THAN (", UNIX_TIMESTAMP(FUTURE_TIMESTAMP), "));");
                -- Run the partitioning query
                PREPARE STMT FROM @__PARTITION_SQL;
                EXECUTE STMT;
                DEALLOCATE PREPARE STMT;
        END IF;END$$DELIMITER ;

将上面4个存储过程语句写到一个sql文件里,partition_call.sql

执行:mysql  -uzabbix -p'zabbix' zabbix < partition_call.sql

使用存储过程:

mysql>  CALL partition_maintenance('<zabbix_db_name>', '<table_name>', <days_to_keep_data>, <hourly_interval>, <num_future_intervals_to_create>)

zabbix_db_name:库名

table_name:表名

days_to_keep_data:保存多少天数据

hourly_interval:每隔多久生成一个分区

num_future_intervals_to_create:本次一共生成多少个分区

例如:

mysql> CALL partition_maintenance('zabbix', 'history', 7, 24, 7);

这个例子就是history表最多保存7天数据,每隔24小时生成一个分区,这次一共生成7个分区

统一调用存储过程:

DELIMITER $$
CREATE PROCEDURE `partition_maintenance_all`(SCHEMA_NAME VARCHAR(32))BEGIN
                CALL partition_maintenance(SCHEMA_NAME, 'history', 28, 24, 14);
                CALL partition_maintenance(SCHEMA_NAME, 'history_log', 28, 24, 14);
                CALL partition_maintenance(SCHEMA_NAME, 'history_str', 28, 24, 14);
                CALL partition_maintenance(SCHEMA_NAME, 'history_text', 28, 24, 14);
                CALL partition_maintenance(SCHEMA_NAME, 'history_uint', 28, 24, 14);
                CALL partition_maintenance(SCHEMA_NAME, 'trends', 730, 24, 14);
                CALL partition_maintenance(SCHEMA_NAME, 'trends_uint', 730, 24, 14);END$$
DELIMITER ;

将这条语句保存成sql文件 partition_all.sql,再次导入存储过程

mysql  -uzabbix -p'zabbix' zabbix < partition_all.sql

计划任务每天调用一次:

注意:

若数据量比较大,首次执行时间较长,请使用nohup执行(我当时执行了15个小时左右,这期间zabbix是无法正常工作,获取agent数据不展示,但数据不会丢失)

nohup time mysql -uzabbix -pzabbix zabbix -e "CALL partition_maintenance_all('zabbix');" &> /tmp/file.txt &

后面只需要调用这个存储过程就可以了,每天执行一次:

mysql -uzabbix -pzabbix zabbix -e "CALL partition_maintenance_all('zabbix');"

写成crontab:

# crontab -e

0 1 * * * /data/tools/mysql/bin/mysql -uzabbix -pzabbix zabbix -e "CALL partition_maintenance_all('zabbix');"

执行脚本:

mysql> CALL partition_maintenance('zabbix', 'history', 28, 24, 14);
+-----------------------------------------------------------+
| msg                                                       |
+-----------------------------------------------------------+
| partition_create(zabbix,history,p201404160000,1397718000) |
+-----------------------------------------------------------+
1 row in set (0.39 sec)
+-----------------------------------------------------------+
| msg                                                       |
+-----------------------------------------------------------+
| partition_create(zabbix,history,p201404170000,1397804400) |
+-----------------------------------------------------------+
1 row in set (0.51 sec)
mysql> CALL partition_maintenance_all('zabbix');
+----------------+--------------------+
| table          | partitions_deleted |
+----------------+--------------------+
| zabbix.history | N/A                |
+----------------+--------------------+
1 row in set (0.01 sec)
....
....
....
+--------------------+--------------------+
| table              | partitions_deleted |
+--------------------+--------------------+
| zabbix.trends_uint | N/A                |
+--------------------+--------------------+
1 row in set (22.85 sec)
Query OK, 0 rows affected, 1 warning (22.85 sec)

[root@hk-zabbix ~]# mysql -uzabbix -p'zabbix' zabbix -e "CALL partition_maintenance_all('zabbix');"

mysql: [Warning] Using a password on the command line interface can be insecure.

+-----------------------------------------------------------+

| msg                                                       |

+-----------------------------------------------------------+

| partition_create(zabbix,history,p201811080000,1541692800) |

+-----------------------------------------------------------+

+-----------------------------------------------------------+

| msg                                                       |

+-----------------------------------------------------------+

| partition_create(zabbix,history,p201811090000,1541779200) |

关闭housekeeping:

zabbix上怎么对mysql数据库做分区表

完成。

“zabbix上怎么对mysql数据库做分区表”的内容就介绍到这里了,感谢大家的阅读。如果想了解更多行业相关的知识可以关注亿速云网站,小编将为大家输出更多高质量的实用文章!

向AI问一下细节

免责声明:本站发布的内容(图片、视频和文字)以原创、转载和分享为主,文章观点不代表本网站立场,如果涉及侵权请联系站长邮箱:is@yisu.com进行举报,并提供相关证据,一经查实,将立刻删除涉嫌侵权内容。

AI