Automate Build Deploy and Process of a datawarehouse

Question

Automate Build Deploy and Process of a datawarehouse

Jim Dillon-291542

Hall of Fame

Points: 3344
More actions
February 28, 2007 at 9:51 am

#119069

Does anyone know how to automate the build, deploy functions of AS2005? I've built a few cubes and am getting tired up starting up SSBIDS, Business Intelligence application, every day to see the new data.
Thanks in advance!
jim

Viewing 6 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic. Login to reply

Jim Dillon-291542 Hall of Fame Points: 3344 More actions · Answer 1

Processing the cube is as easy as an XMLA query, XML for analysis query. Check out this example.

Please rember to change the { to (I can't figure out how to escape them so they'll show correctly in the forum. This is due to my ignorance I'm sure.)

{Process xmlns="http://schemas.microsoft.com/analysisservices/2003/engine"}

{Object}

{DatabaseID}DataWarehouse{/DatabaseID}

{/Object}

{Type}ProcessFull{/Type}

{WriteBackTableCreation}UseExisting{/WriteBackTableCreation}

{/Process}

{/xmla}

I have absolutely no idea of what the url is.

The {DatabaseID} is the name of the data warehouse to process

The {Type} tells the server to process the data warehouse fully

The {WriteBackTableCreation} tells the server to write over the existing data warehouse database{/WriteBackTableCreation}

Basically this example will fully process the data warehouse DataWarehouse and overwrite the existing one.

stevefromOZ SSC-Forever Points: 43646 More actions · Answer 2

Note also that SSIS has quite good suport for reprocessing AS objects too.

reference is not quite true - doesn't this refer to how you want to use/create the WriteBack Table that is the target of cube writebacks? Yes, what's you've described will overwrite (ie process) the entire DB but this is a result of the ProcessFull attribute not the WriteBackTableCreation. This attribute does have some other neat applications - if you modify your Aggregation Design, you can then call a Process of the partition(s) but set the Type to be Index (could be IndexesOnly?) and this will recreate the aggregations without having to re-ead the entire partition rowsource again - very useful when each partition has several million rows.

Steve.

Jim Dillon-291542 Hall of Fame Points: 3344 More actions · Answer 3

Do you have any idea what the url specifies? I can't seem to find any documentation on this.

Thanks,

jim

stevefromOZ SSC-Forever Points: 43646 More actions · Answer 4

Something yet nothing

Because the document is XML it refers to a namespace (which can be used by the app [Analysis Services] to ensure that the document is one that it can use). THe namespace dictates the elements and attributes (tags and properties ) that can exist in the document. A much better explanation of XML Namespaces can be found here.

Personally, i've found quite often that the namespace is somewhat 'logical' versus physical - ie trying to navigate to that URL will not lead you to a XSD (ie an xml schema document). But, disecting the parts of the path can tell you what 'system' the schema is used by, that is, if I mailed you the text you posted originally as an xml document named 'bob.xml', looking at the namespace you coukld tell that really it's a Microsoft schema, most likely used by analysis services, in particular the engine, probably the 2003 version (yes, we all thought in Feb '03 that SQL2K5, or Yukon as it was known then, would be released in 03... ahh, the dreams of youth).

Just for kicks, if you're interested, nearly every function (every function mabe?) that you can perform through SQL Mgmt Studio on your AS databases (and serve rin general) can be XMLA scripted - rather than hitting the 'OK' button next time, look for the 'Script' button. A hint on this, the Script button often looks like it has a submenu, which you can then select to script to clipboard, new file, etc. Simply clicking the button itself will default to scripting to a new xmla script window (saving you one click per time ).

Steve.

Jim Dillon-291542 Hall of Fame Points: 3344 More actions · Answer 5

Thank you very much. That explained the url very nicely.

jim