Ingeniero

DSVs are used just like Connection Managers. However, there are a few key differences to remember
when using them. Like data sources, DSVs allow you to define the connection logic once and reuse it
across your SSIS packages. However, unlike connections, DSV structures are stored once, and then are
disconnected from the real source. This means that the underlying structure of the DSVs maychange,
but the DSVs are not automatically refreshed. This can cause some problems during execution; if you
were to change the Employee table in a connection to a DSV for Human Resources, the DSV would not
pick up the change. On the other hand, if your model does not change often, this type of caching is a
huge benefit in your development of SSIS packages. The DSVs provide the disconnectedcapabilities to
allow development of SSIS packages to continue against cached metadata. DSVs also provide a side
advantage in speeding up package development. Because DSVs are most likely defined as a subset of
the actual Data Source, your SSIS connection dialog boxes will connect, realize data, and subsequently
load much faster.
Precedence Constraints
Precedence constraints direct the tasks toexecute in a given order. In fact, precedence constraints are the
connectors that not only link tasks together but define the workflow of your SSIS package. Constraints
control the execution of the two linked tasks by executing the destination task based upon the final state
of the prior task and business rules that are defined using special expressions. The expression language
embedded in SSISessentially replaces the need to control workflow using script - based methodologies
that enabled and disabled tasks, as was used in the DTS legacy solution. With expressions, you can direct
the workflow of your SSIS package based on all manner of given conditions. We ’ ll go into many
examples of using these constraints throughout this book.
To set up a precedence constraint between two tasks,the constraint value must be set, and optionally
you can set an expression. The next sections give a quick overview of the differences between the two.
Constraint Value
Constraint values define how the package will react when the prior task of two linked tasks completes
an execution. The choices define whether the destination task of two linked tasks should execute based
solely on how theprior task completes. There are three types of constraint values:
Success: A task that ’ s chained to another task with this constraint will execute only if the prior
task completes successfully. These precedence constraints are colored green.
Completion: A task that ’ s chained to another task with this constraint will execute if the prior
task completes. Whether the prior task succeeds or failsis inconsequential. These precedence
constraints are colored blue.
Failure: A task that ’ s chained to another task with this constraint will execute only if the prior
task fails to complete. This type of constraint is usually used to notify an operator of a failed
event or write bad records to an exception queue. These precedence constraints are colored red.
Conditional Expressions
Theconditional expression options that you can apply to a precedence constraint allow you to mix in a
dynamically realized expression with the constraint value to determine the package workflow between
two or more linked tasks. An expression allows you to evaluate whether certain conditions have been met
❑
❑
❑
Chapter 1: Welcome to SQL Server Integration Services
12
before the task is executedand the path followed. The constraint evaluates only the success or failure of
the previous task to determine whether the next step will be executed. The SSIS developer can set the
conditions by using evaluation operators. Once you create a precedence constraint, you can set the
Evaluation Option property to any one of the following options:
Constraint: This is the default setting and...