Print all column names in SSMS (SSMS trick)

Every once in a while I come across some “hidden” and nice features in SSMS. One of them is really nice if you want to print all columns names of a table.

So instead of writing all the column names by hand you could follow these easy steps.

– Open SSMS
– Expand you database
– Expand your table

expand-ssms

And then drag the folder called “Columns” to your query window

all-columns

Advertisements

How to make your model smaller

One off the biggest mistakes out there is to add columns with unique values to a model. Such as order numbers and orderline numbers. Unless you have a very good reason to do so. This will make your models larger then necessary. And the SSAS engine will not aggregate your data as good as it should.

Lets make a sample. We will use AdventureWorksDW2012 as we always do.

First I will add a couple of columns to my fact

alter table [dbo].[FactInternetSales]
add MostLikelyUniqueVarchar varchar(10),
UniqueIntValue int IDENTITY(1,1)

update [dbo].[FactInternetSales]
set MostLikelyUniqueVarchar = SUBSTRING(CONVERT(varchar(255), NEWID()), 0, 10)

Then I will create a tabular model with the following tables

– DimDate
– FactInternetSales
– DimProduct

And it will look something like this

Tables

I mark DimDate as date table and then deploy my model to the server.

The model is now about 6 MB.

Size 1

Then I replace my FactInternetSales with this query. Still having SalesOrderNumber and SalesOrderLineNumber in it.

select
      ,[OrderDateKey]
      ,[DueDateKey]
      ,[ShipDateKey]
      ,[CustomerKey]
      ,[PromotionKey]
      ,[CurrencyKey]
      ,[SalesTerritoryKey]
      ,[SalesOrderNumber]
      ,[SalesOrderLineNumber]
      ,[RevisionNumber]
      ,[OrderQuantity]
      ,[UnitPrice]
      ,[ExtendedAmount]
      ,[UnitPriceDiscountPct]
      ,[DiscountAmount]
      ,[ProductStandardCost]
      ,[TotalProductCost]
      ,[SalesAmount]
      ,[TaxAmt]
      ,[Freight]
      ,[CarrierTrackingNumber]
      ,[CustomerPONumber]
      ,[OrderDate]
      ,[DueDate]
      ,[ShipDate]
  FROM [dbo].[FactInternetSales]

The model is then down to about 4 MB.

Size 2

But what If I remove all the unique columns? Such as CarrierTrackingNumber, CustomerPONumber, SalesOrderNumber and SalesOrderLineNumber.

SELECT [ProductKey]
      ,[OrderDateKey]
      ,[DueDateKey]
      ,[ShipDateKey]
      ,[CustomerKey]
      ,[PromotionKey]
      ,[CurrencyKey]
      ,[SalesTerritoryKey]
      ,[OrderQuantity]
      ,[UnitPrice]
      ,[ExtendedAmount]
      ,[UnitPriceDiscountPct]
      ,[DiscountAmount]
      ,[ProductStandardCost]
      ,[TotalProductCost]
      ,[SalesAmount]
      ,[TaxAmt]
      ,[Freight]
      ,[OrderDate]
      ,[DueDate]
      ,[ShipDate]
  FROM [dbo].[FactInternetSales]

Now the model is reduced to about 3.5 MB. Half the size we began with 🙂

Size 3

So the conclusion is. Do not add columns with unique values in your model. You should only have columns with foreign keys and measures.

Optimize your datatypes for SSAS

Not everyone is aware of the fact that not all data types are supported in SSAS cubes / models. If you use one of the data types that are not supported by SSAS, SSAS will cast all your unsupported data types to supported data types when the cube is being processed.

This will cause your cube processing time to increase. And you will use more CPU than necessary.

Microsoft has provided a list of supported data types here

To avoid these kinds of problems there are several things you can do.

  • Use views as source for your cubes. Then cast the values in these views. SQL Server is better than SSAS to cast values. And it is more likely that your SQL Server has more power that your SSAS server.
  • Make sure your DWH only uses supported SSAS data types. But that’s not to likely is it?