I wasn't quite sure where to post this message but I guess we can start off in this Forum. I've been reading the DSXChange newsletter and came across the
Server to Parallel Transition Lab
article.
It's the point about the next release of DataStage (Hawk) that is going to be released that interests me and hope that it is just my misunderstanding of the article.
It states
The next release has been touted as paradigm change since it is the first major release of DataStage to be solely based on the parallel (PX) engine (also called DataStage Enterprise Edition).
What does this actually mean in terms of the effects of DataStage Server ? Will Server Jobs disappear ?
There are only 10 kinds of people in the world, those that understand binary and those that don't.
The info you were referring to in the newsletter has been corrected (take a look at it now). We apologize for the confusion it caused.
However, in general, let me state a couple of personal observations from interactions with Ascential / IBM. They are heavily promoting Enterprise Edition (PX) in all of their conferences and literature. You will see limited references in which they promote the Server edition.
Now considering that the majority of their user-base is on Server, you might ask "Why?". My understanding is that they feel there are differences in the architecture of the EE product and its parallel capabilities that will provide both them (from a maintenance / enhancement standpoint) and their customers (from a features standpoint) with advantages as they head into the future. So they are heavily promoting the "newer" product to slowly transition their customer base.
Please don't interpret this to mean that Server is "going away". This just means that over time more and more sites will be EE-based instead of Server-based. This DOES mean that we've noticed an increase in posts from people requesting "PX" or "EE" DataStage developers, a much rarer skill set right now. Many of these requests are coming from IBM's largest customers, who are using the parallel capabilities inherent in the new product to achieve some staggering throughputs.
As to the technical aspects of getting your Server-based jobs into an EE-based environment, I believe they must be re-written due to the fact that the components used to build the jobs are different (though the interface and much of the methodology is similar). However, I'll let one of our technical wizards like Ray or Craig post more on that!
Thanks for the response - looks like we won't be going towards Hawk for a while until I find the time to learn PX.
I'd love to come out to the conference but seeing as I have never even had a local based training course I think the chances are very remote but I'll keep asking !!.
Looks like I'll have to go through the same painful learning curve for PX that I've done for Server Jobs - on my own but thankfully with a forum full of excellent articles/help.
There are only 10 kinds of people in the world, those that understand binary and those that don't.
I haven't seen the Hawk release yet but I would expect Server Edition Hawk to be an improvement over Server Edition 7.x. You get all the improvements in the Designer such as better metadata reporting, better impact analysis, better developer collaboration and improved job sequences.
Ascential have also made efforts to make EE look and act more like Server Edition when they put the Lookup and Transformer GUI of server jobs into parallel jobs. When they bring in the SCD stages and tidy up some of the more difficult parallel stages like Modify I would expect it to become easier to use.
I would like to see Server Edition become the equivelent of DB2 Express, a cheaper version for small to medium businesses that keeps it simple.
