70-463 · Question #224
You are performance tuning a SQL Server Integration Services (SSIS) package to load sales data from a source system into a data warehouse that is hosted on Windows Azure SQL Database. The package cont
The correct answer is A. Set the EngineThreads property of the data flow task to 7.. The EngineThreads property on the Data Flow task controls how many threads the SSIS data flow engine allocates to process execution trees in parallel. Each execution tree is a segment of the data pipeline that can run on its own thread. The default value of EngineThreads is typic
Question
You are performance tuning a SQL Server Integration Services (SSIS) package to load sales data from a source system into a data warehouse that is hosted on Windows Azure SQL Database. The package contains a data flow task that has seven source-to-destination execution trees. Only three of the source-to-destination execution trees are running in parallel. You need to ensure that all the execution trees run in parallel. What should you do?
Options
- ASet the EngineThreads property of the data flow task to 7.
- BSet the MaxConcurrentExcecutables property of the package to 7.
- CCreate seven data flow tasks that contain one source-to-destination execution tree each.
- DPlace the data flow task in a For Loop container that is configured to execute seven times.
How the community answered
(40 responses)- A60% (24)
- B10% (4)
- C25% (10)
- D5% (2)
Explanation
The EngineThreads property on the Data Flow task controls how many threads the SSIS data flow engine allocates to process execution trees in parallel. Each execution tree is a segment of the data pipeline that can run on its own thread. The default value of EngineThreads is typically 5, which is why only 3 of the 7 trees were running in parallel (the engine was thread-limited). Setting EngineThreads to 7 gives the engine enough threads to run all 7 execution trees simultaneously. Option B (MaxConcurrentExecutables) controls how many control flow tasks run in parallel across the package, not execution trees within a single data flow task. Options C and D are workarounds that restructure the package unnecessarily.
Topics
Community Discussion
No community discussion yet for this question.