Chapter 14
Reliability and Scalability
WHAT’S IN THIS CHAPTER?
- Restarting packages
- Using package transactions for data consistency
- Using error outputs for reliability and scalability
- Scaling out efficiently and reliably
Reliability and scalability are goals for all your systems, and although it may seem strange to combine them in one chapter, they are often directly linked, as you will see. Errors and the unexpected conditions that precipitate them are the most obvious threats to a reliable process. Several features of SQL Server 2012 Integration Services enable you to handle these situations with grace and integrity, keeping the data moving and the systems running.
Error outputs and checkpoints are the two features you will focus on in this chapter, and you will see how they can be used in the context of reliability. Implementation of these methods can also have a direct effect on package performance, and therefore scalability, and you will learn how to take into account these considerations for your package and process design. The capability to provide checkpoints does not natively extend inside the Data Flow, but there are methods you can apply to achieve this. The methods can then be transferred almost directly into the context of scalability, enabling you to partition packages and improve both reliability and scalability at the same time. All of these methods can be combined, and while there is no one-size-fits-all solution, this chapter describes all your options so you ...
