deployment method
In Redshift
1, Deploy process: Drop and Refresh the view, Drop table, Create an empty table using the DDL
2, Build data: Insert the data to table from the view
In Netezza, when do deployment:
1, Deploy process: Drop all the existing tables, Refresh the view and create the an empty table for that view
2, Build data: Create the table with data(CTAS from view), replace the empty table with the new table
When building data, the scripts will first create an table with an incoming suffix, then do some data validation, if all passed, then replace the empty target table with the incoming table.
Jenkins slaves is an EC2 Linux server.
In the slave, it calls shell script to run the SQL(refresh views and tables)
It also run shells which connect to the DMA server to run the DMA jobs
Using pipeline to get Jenkins jobs connected
continuous integration and continuous deployment in DW/BI
原文：http://www.cnblogs.com/davablog/p/6266476.html