Server to Parallel job conversion

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
atul9806
Participant
Posts: 96
Joined: Tue Mar 06, 2012 6:12 am
Location: Pune
Contact:

Server to Parallel job conversion

Post by atul9806 »

Hi All
I know there is no way to convert the Server job to Parallel job in a single click, we have to redesign the job again. But I want your suggestions in this task ?
What are the things for which I have to consider/take care ?
~Atul Singh
<a href=http://www.datagenx.net>DataGenX</a> | <a href=https://www.linkedin.com/in/atulsinghds>LinkedIn</a>
rohitagarwal15
Participant
Posts: 102
Joined: Thu Sep 17, 2009 1:23 am

Re: Sever to Parallel job conversion

Post by rohitagarwal15 »

You can create server shared container and can use that in your parallel job if this matches your requirement otherwise go for redesigning the job in parallel but some of server job stages are not available in parallel job so you need to tke care of that.
Rohit
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

The first thing to consider is whether it makes sense to undertake the "conversion". Server jobs are not going away. And they are more efficient for small data volumes than are parallel jobs.

Other than that, the best way to convert is, indeed, to analyze the logic of the server job and to design a parallel job from the ground up to effect the same logic.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
oracledba
Premium Member
Premium Member
Posts: 49
Joined: Mon Aug 06, 2012 9:21 am

Post by oracledba »

Some things to consider when doing Sever to Parallel job conversion


-Datastage parallel jobs can run in parallel on multiple nodes. Server jobs do not run on multiple node.
-Parallel jobs support partition parallelism(Round robin,Hash,modulus etc.), server jobs don't support this.
-The transformer in Parallel jobs compiles in C++. In server jobs, the transformer is compiled in Basic language.
-Transformer functions are few in parallel jobs vs in server jobs more Transformer functions available
-In parallel jobs look up stage is available. In server jobs no look up stage is available so you need to do the logic using a transformer which serves as a look up in server jobs
-server job runs on one node whereas parallel job runs on more than one node.
atul9806
Participant
Posts: 96
Joined: Tue Mar 06, 2012 6:12 am
Location: Pune
Contact:

Post by atul9806 »

Thanks man
Any more suggestions guys...
~Atul Singh
<a href=http://www.datagenx.net>DataGenX</a> | <a href=https://www.linkedin.com/in/atulsinghds>LinkedIn</a>
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

I'm just going to echo what Ray said but add that you should think of this as if you were converting from a completely different tool to another, which basically you are. Make sure you thoroughly understand what the Server job does and then build an equivalent Parallel job. Much the same as if you'd started from an Informatica or Pentaho job... emulate what the old tool did but leverage the strengths of the new tool.

I don't feel there's anything special to "consider / take care of" during this, just all of the normal things you have to consider when building a Parallel job any day of the week.
-craig

"You can never have too many knives" -- Logan Nine Fingers
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

Or, there are several organizations who will do this for you for a fee. And you probably have to agree to your job designs being offshored.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Post Reply