DSXchange

At this point in time you need to define your column names and data types when you develop the job. If your source is a database table then the expectation is the metadata on the table does not change, if it does change the DataStage job needs to be changed. If your source is a text file or your dat...

Backwards functionality! There have been a couple handy features in 7.0 that are no longer in 7.5. One of the reasons why the Hawk beta testing will be useful.

There is the Index function which finds the starting character position of a substring. It can be used in a parallel transformer. Not sure if it will work in a filter stage. Constraint: Index(input.mycol, "MI") > 0

I would filter them out in the initial SQL select if you don't want them involved in your job in any way, I would filter them out in the job via a constraint if you don't want to process them but you want to write them to a reject file for data quality investigation.

There is nothing wrong with the surrogate key stage. Whatever method you use if you are going to generate keys within the job you need to provide the job with the last high value used via a job parameter. Then you can use the surrogate key stage which is the specialised stage for this task and will ...

Is any one using a good range lookup custom stage? There was talk of one being available via Ascential services. Currently range lookups are messy in parallel jobs and it would be good to have a more efficient option.

A direct mailing company, Pitney Bowes, has bought First Logic! First Logic is one of the biggest DQ vendors in the US and has OEM agreements with Business Objects and Informatica (and Ab Initio?). Informatica may have seen this coming as they signed a new OEM with Trillium. Pitney Bowes also bought...

In parallel jobs there is a GetEnvironment function but I don't know of an equivelent one in BASIC. You could you DSExecute to run an echo command. It's not a very good programming practice as it stops your routine from operating as a black box. It can lead to misunderstandings from support in produ...

There is a FAQ on this topic. You can read it at:
http://www.dsxchange.com/viewtopic.php?t=92038

Or you can view the modifiable version at:
http://wiki.ittoolbox.com/index.php/HOW ... _variables

Why don't you round the value off and match it exactly? In a transformer prior to the lookup create a stage variable that rounds to the nearest digit. If the resulting value is greater then the original value then reduce it by 0.5. This gives you rounding to the nearest 0.5. If this value is > 0 and...

Agree, since you already plan to have a transformer in your job I would do it all in that transformer. The FIELD command should be able to pick the columns out of your varchars quite easily and you may need stage variables or you may be able to do it straight in the derivation fields. The FIELD func...

Run Export project, check the "Export as XML document" tab and then go to the XML tab. You will see a "Style sheet" option. Make sure this is checked and set to the default DSReport-Waterfall.xsl in your DataStage program directory. Run this export. Open up the XML document in a web browser, you wil...

I keep a test job around that has every data type as input and output. I use it to test parallel transformation and modify functions such as these. It lets me output the final result and interim values such as the value produced by DecimaltoString. You may need to add formatting to your DecimaltoStr...

I didn't say there would be a demise of server jobs, in fact I think they will be supported for years to come. I said there would be no new functionality in server jobs. There will be bug fixes in server jobs because there are a lot of people using them. Server jobs may also benefit from some enhanc...

I wouldn't expect any new functionality for server jobs from here on. Anything in the new release such as wizards, templates and a scd specialised stage will be developed for parallel jobs.