Server Architecture and Command Execution - Microsoft SQL Server 2008 Analysis Services - page 497

<row>
<Object>
<Object>
<MeasureGroupID>Warehouse Inventory</MeasureGroupID>
<CubeID>FoodMart2008</CubeID>
<DatabaseID>FoodMart 2008</DatabaseID>
</Object>
</Object>
<Impact>Unprocessed</Impact>
<Severity>Warning</Severity>
<Faultcode>-1056964608</Faultcode>
<Description>Object will be unprocessed/cleared.</Description>
</row>

...

</root>
</results>
</return>

Inside the batch, you can group commands in blocks. Block commands are a useful mech-
anism that helps optimize the server's performance when doing complex processing oper-
ations because it allows you to define the order in which commands execute. To maximize
capabilities of the server, you can bind commands inside the block that will execute in
parallel, by using the <Parallel> element, as shown in Listing 26.14.

LISTING 26.14

The Syntax of the Parallel Execution Block

<Parallel maxParallel=n >
<Command1>cmd_1</Command1>
<Command2>cmd_2</Command2>
<Command3>cmd_3</Command3>
</Parallel>

With this command, you tell the server to execute commands from the list in parallel. If
you define the maxParallel parameter on the Parallel block, Analysis Services tries to
execute as many commands in parallel as specified by the maxParallel parameter. If
maxParallel is not specified, the server tries to execute all the commands in the block at
the same time. After all commands in the parallel block are completed, the command next
to the block starts to execute. When executing the commands, Analysis Services estimates
the resources it has, the resources required for execution, and then decides how many
commands it can execute in parallel. Therefore, the maxParallel property is only a hint
for the server.