数据牛刀是一款提供大数据表自动分割、归档和清理过期数据的服务。它使用C# /.NET CORE实现,当前支持Mysql数据源。
Data Niu Knife is a service provides to automatically splits, archive, and clean up stale data for big data tables. It is implemented using C# /.NET CORE and currently supports Mysql data sources.
核心功能:
- 通过简单的配置,可管理任意多个数据库节点上的大数据表的定期数据分割、归档和清理。
- 每日执行数据复制。
- 按指定天数移除过期数据。
- 按月归档数据。
Key Features:
- through simple configuration, can manage the regular data segmentation, archiving and cleaning of big data tables on any number of database nodes.
- Perform data replication daily.
- Remove expired data by the specified number of days.
- Archive data by month.
希望使用此服务的大数据表,有如下要求:
- 数据表存在自增长ID;
- 数据表存在时间列,数据以时间顺序增长;
- 数据表不存在外键约束;
- 提供一个拥有读取数据表元数据和建表权限的数据库账号;
If you want to use this service be sure following requirements:
- the big data table is based on the self-growth ID;
- the big data table has time column, the data grows by time order;
- the big data table does not have a foreign key at all;
- need a DB Account with the authority to read the data table metadata and create the data table;
配置示例: Configuration Sample:
{
"AppSettings": {
"LogManPath": "/app/LogMan/"
},
"MySqlClusterSettings": {
"Nodes": [
{
{
"MysqlNode": {
"ID":1,
"IsSlave": false,
"DataBasesName": "data_sharding_a",
"ConnStr": "server=192.168.3.250;database=data_sharding_a;user=app_user;password=your pwd;charset=utf8;",
"DevideFromNodeID": 0,
"DevideDataSet": "table 1:hash key,table 2:hash key,table n:hash key",
"AutoMoveDataSet": "table_name=data_shard,key_name=id,date_field=created,data_hold_days=180,archive_node_id=2,schedule_time=23:00:00;"
}
},
{
"MysqlNode": {
"ID": 2,
"IsSlave": false,
"DataBasesName": "data_sharding_b",
"ConnStr": "server=192.168.3.250;database=data_sharding_b;user=app_user;password=your pwd;charset=utf8;",
"DevideFromNodeID": 0,
"DevideDataSet": "",
"AutoMoveDataSet": ""
}
}
]
},
"AllowedHosts": "*"
}
理解这一段配置: Understand this paragraph configuration:
"AutoMoveDataSet": "table_name=data_shard,key_name=id,date_field=created,data_hold_days=180,archive_node_id=2,schedule_time=23:00:00;"
table_name: 要列入自动管理的数据表名。 the name of data table that you want to use this service.
key_name: 主键名。 main key on the data table.
date_field: 时间列名。 the date time column.
data_hod_days: 数据保留期的天数。 how many days for keep the data before removed.
archive_node_id: 本配置中作为归档库的数据库节点ID。 the data archive DB node id in this configures.
schedule_time: 每日运行的计划时间。 the time for start jobs per day.
本服务可以在docker容器中运行: This service supports docker container:
docker build -f Dockfile -t data_niu_knife:demo .
docker run --name data_niu_knife_hosted --mount type=bind,source=/home/docker_data/DataNiuKnife/LogMan/,target=/app/LogMan/ -d data_niu_knife:demo .
正常运行后可看到类似下方的日志: After start the service, you can see some logs like below:
[root@localhost ~]# cat /home/docker_data/DataNiuKnife/LogMan/MysqlDataWorker-2019-07-29-pid1.log
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:00 ======
MysqlDataWorker已在服务环境启动,源表名:data_shard
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:00 ======
开始自动创建分表,表名:data_shard_spt_201907
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:00 ======
开始创建分表:
CREATE TABLE IF NOT EXISTS `data_shard_spt_201907` (
`id` bigint(10) unsigned NOT NULL AUTO_INCREMENT,
`value` varchar(255) DEFAULT NULL,
`created` datetime NOT NULL,
PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=51 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_0900_ai_ci
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:01 ======
成功创建分表:data_shard_spt_201907
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:01 ======
成功自动创建分表,表名:data_shard_spt_201907
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:01 ======
开始自动复制数据,源表名:data_shard,分表名:data_shard_spt_201907
====== DataNiuKnife.MysqlDataWorker Info 07/29/2019 11:00:02 ======
自动复制了50笔数据,源表名:data_shard,分表名:data_shard_spt_201907