THE SQL Server Blog Spot on the Web

Welcome to SQLblog.com - The SQL Server blog spot on the web Sign in | |
in Search

SQLBI - Marco Russo

SQLBI is a blog dedicated to building Business Intelligence solutions with SQL Server.
You can follow me on Twitter: @marcorus

Parallelize incremental processing in Tabular #ssas #tabular

I recently came in a problem trying to improve the parallelism of Tabular processing. As you know, multiple tables can be processed in parallel, whereas the processing of several partitions within the same table cannot be parallelized. When you perform an incremental update by adding only new rows to existing table, what you really do is adding rows to a partition, so adding rows to many tables means adding rows to several partitions. The particular condition you have in this case is that every partition in which you add rows belongs to a different table.

Adding rows implies using the ProcessAdd command; its QueryBinding parameter specifies a SQL syntax to read new rows, otherwise the original query specified for the partition will be used, and it could generate duplicated data if you don’t have a dynamic behavior on the SQL side. If you create the required XMLA code manually, you will find that the QueryBinding node that should be part of the ProcessAdd command has to be moved out from ProcessAdd in case you are using a Batch command with more than one Process command (which is the reason why you want to use a single batch: run multiple process operations in parallel!).

If you use AMO (Analysis Management Objects) you will find that this combination is not supported, even if you don’t have a syntax error compiling the code, but you might obtain this error at execution time:

The syntax for the 'Process' command is incorrect. The 'Bindings' keyword cannot appear under a 'Process' command if the 'Process' command is a part of a 'Batch' command and there are more than one 'Process' commands in the 'Batch' or the 'Batch' command contains any out of line related information. In this case, the 'Bindings' keyword should be a part of the 'Batch' command only.

If this is happening to you, the best solution I’ve found is manipulating the XMLA code generated by AMO moving the Binding nodes in the right place. A more detailed description of the issue and the code required to send a correct XMLA batch to Analysis Services is available in my article Parallelize ProcessAdd with AMO. By the way, the same technique (and code) can be used also if you have the same problem in a Multidimensional model.

Published Tuesday, September 11, 2012 3:30 PM by Marco Russo (SQLBI)

Comment Notification

If you would like to receive an email when updates are made to this post, please register here

Subscribe to this post's comments using RSS

Comments

No Comments

Leave a Comment

(required) 
(required) 
Submit

About Marco Russo (SQLBI)

Marco Russo is a consultant, writer and trainer specialized in Business Intelligence with Microsoft technologies. He runs the SQLBI.COM website, which is dedicated to distribute resources useful for BI developers, like Integration Services components, Analysis Services models, tools, technical information and so on. Marco is certified as MCT, MCDBA, MCSD.NET, MCSA, MCSE+I.

This Blog

Syndication

Archives

Powered by Community Server (Commercial Edition), by Telligent Systems
  Privacy Statement