Chinaunix首页 | 论坛 | 博客
  • 博客访问: 62606
  • 博文数量: 33
  • 博客积分: 0
  • 博客等级: 民兵
  • 技术积分: 345
  • 用 户 组: 普通用户
  • 注册时间: 2014-08-02 08:41
文章分类

全部博文(33)

文章存档

2015年(13)

2014年(20)

我的朋友

分类: 数据库开发技术

2015-04-16 08:38:10

Sometime a user wants to add a shard (i.e., go from 3 shards to 4). What is the process? Clients need to re-connect to all 4 shards — close the DistibutedEngine that is connected to 3 shards and reconnect to 4. The moving/rebalancing of the data is the application’s responsibility. There is no right way of redistributing data across shards this at the moment. Essentially its necessary to pull data to the client and redistribute from the client

The first approach is to use a file to read the data to:

  • Start xsqlcmd

点击(此处)折叠或打开

  1. XSQLCMD> format CSV
  2. XSQLCMD> output mytable.csv
  3. SELECT * FROM MYTABLE


  • Copy mytable.csv to all shards and

    点击(此处)折叠或打开

    1. INSERT INTO MYTABLE from select * from foreign table (path='mytable.csv') where ...

阅读(493) | 评论(0) | 转发(0) |
0

上一篇:在eXtremeDB中创建相同的表

下一篇:没有了

给主人留下些什么吧!~~