Some time ago, I tried to port the sqlserver database to the mysql database.
The solution is that the perl language reads sqlserver for corresponding type conversion data processing and writes it to mysql
Two executors and one receiver
The final import speed is 10 million-level tables, which are completed in about 5 minutes. The garbled code and other problems are solved.
3G databases take half an hour
15g takes 5 hours
The process is as follows:
The first is environment configuration and plug-in installation.
First, the acceptor
In a linux system
Install mysql
If the geographic coordinate field geo exists, mysql must use a version later than 5.5.
CentOS6.4 install mysql5.6.14
Then, configure the execution terminal (you can install it on the host of another linux system or on the host of mysql)
Here we will divide the database tables into several parts to run the data. Therefore, if the bottleneck of SQL Server reading and network speed is reached, the faster the execution end is.
Install perl on CentOS
Install DBDmysql in Perl to operate mysql Databases
Perl connects to SQLServer through unixODBC
Finally, install handlersocket (because handlersocket has installation steps on both the acceptor and the acceptor)
HandlerSocket introduction and installation and uninstallation