Decision Stream 7.1 Command Line argument to run Decision Stream Build 1

[hiral265]

Can anyone tell me the Command Line argument to run Decision Stream Builds? In my SQL Server Enterprise Manager I have a job where I need to give a command to run a build.

I know the command to run Decision Stream Job Streams but I am not sure how to run builds directly. I want to avoid creating a Job Stream and put builds there to run them.

Thanks,
Heeral

 

[mfgf]

Hi,

A jobstream is fired off using the rundsjob.exe command. Fact builds use databuild.exe, and dimension builds use dimbuild.exe. If you want to find out what the relevant command is to fire off a build, open the DecisionStream catalog in the designer, right-click on the relevant build, and select 'execute'. This will take you to an execute dialog window, and near the bottom you will see the command to launch the build - you can even select and copy the command to the clipboard from here. Just be careful when you paste it in to Enterprise Manager - make sure you remove the -P switch which gets the build to pause on completion until the return key is pressed.

Regards,

MF.

 

[hiral265]

Hi MFGF,

Thanks for the help. It works...

Just one more curious question. rundsjob.exe is for jobstreams and what is rundsnode.exe for?

Thanks,
Heeral

 

[mfgf]

Hi,

Within a jobstream, most of the nodes will run as part of the jobstream executable (rundsjob.exe). This applies to SQL nodes, procedure nodes etc. The exceptions are the build nodes, which each fire off separate executables (databuild.exe and dimbuild.exe). If you want your jobstream to run two dimension builds in parallel on a 2-cpu machine, this is ideal as each will be be spawned as a separate executable and can be allocated to a separate CPU. If however you want your jobstream to run (say) two SQL nodes in parallel, normally these will run as part of the main executable for the jobstream (rundsjob.exe) and so will use the same CPU. You can change this behaviour for these nodes by checking the 'execute in a separate process' checkbox in the properties of each node. This then forces the jobstream to spawn separate rundsnode.exe processes for the nodes, and these can then be allocated to separate CPUs by the OS.

Hope that didn't bore you too much!

Best regards,

MF.