一张大表的初始同步，从前提到集线器，前提CPU固定为99％ [英] initial sync of one large table, premise to hub, premise CPU pegged at 99%

问题描述

 我们有一个初始同步组设置，用于从场所到天青的270万条记录表.我们将DTU值提高到1600，以帮助同步运行更快.  

 We have an initial sync group setup for a 2.7 million record table from a premise to our azure. We upped the DTU thing to 1600 to help the sync run faster.  

前提服务器具有8G RAM，而不是一吨的存储量，它是带有ms sql server 2014的2016服务器.该前提服务器上的CPU同步运行时固定为99％.

The premise server has 8G RAM, not a ton of storage, its a 2016 server with ms sql server 2014. The CPU at this premise server pegs at 99% while the sync is running.

在天蓝色的一面是否有任何要调整或配置的内容，这样它就不会以这种方式占用客户驻地服务器以进行初始数据提取.即使一个小时左右，这似乎也是无法接受的.前提是服务器当然是用于其他 事情！

Is there anything to tweak or configure on the azure side so that it doesn't tie up the customers premise server in this way for an initial data pull. It seems unacceptable even for an hour or so. The server at the premise is of course used for other things!  

推荐答案

Hello,

Data Sync won't know the two databases are identical until it compares the data row by row. It is a very costly process and may take a long time and a lot of resources if you have large databases/tables. Our recommendation is to have data only on one side and keep the same table empty in the other databases. In this case, data sync will use bulk load during initialization and it is much faster than row by row comparison.

Regards,

Alberto Morillo
SQLCoffee.com

这篇关于一张大表的初始同步，从前提到集线器，前提CPU固定为99％的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！