I'ved used PostgreSQL for several important databases. Although I prefer SQL Server for many reasons, PG works well in a Unix environment. It even has a few interesting...

I'm not against using a view, just not yet. I think it's a very interesting idea. I'm not a big fan of views (prefering stored procedures), but in...

Steve,

Each job appears only once in the fact, as that is that cube's level of granularity, so it's not a bad assumption.

I have never based a cube on a view...

dtbl is a dimension table. So the problem is that I have a value in the dimension that I want to use in a fact table calculation. The...

Baddog,

Would you mind providing more specifics for how you made that work? I'm trying to do just what you did, and am also having issues.

Thanks,

Here is a much reduced version of the problem with the SQL solution.

CREATE TABLE dtblCalendar (

intDateID int NOT NULL PRIMARY KEY,

dtmDate smalldatetime NOT NULL,

intDaysPassed smallint NULL

);

GO

CREATE TABLE ftblJob...

You can hook the fail arrow of the destination to a text file or a table and set the failure to redirect row. This will allow the good rows...

Wanted to go back to the issue of the columns to PK themselves. Sometimes the answer is to only PK the intersecting columns. You do this when you...

This answer to this question is, like the answer to so many questions, is it depends. At least in part. Truncating and reloading can be used to get...

I see you posted this a long time ago. I've only run across it while researching an issue I am having. I assume that you are using 1...

Do you know for sure the DSNs were set up as System DSNs?

I wouldn't rule out a transaction based fact table, as you've indicated a change would cause a row insertion. If it's your first cube, you've chosen a very challenging...

The MS build-in SCD component is for example not that efficient. It uses the OLE DB Command to perform the updates, which means that if a million rows have to...

Why would you never use it? I'm not a big fan of most wizards, but I find this one very effective.

Sometimes the database has so many design issues that the final effect of performance tuning is going to be limited. In this case a second database is likely worth...