PX Retraining times?

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
roy
Participant
Posts: 2598
Joined: Wed Jul 30, 2003 2:05 am
Location: Israel

Post by roy »

Hi,
I'm facing the same situation.
IMHO only:
As I see it you need to go thru the concept change of how things are done, go over the different stages which are mostly a new way to do things you normaly do in the server edition.
on top of that the 7.5 has the advpx.pdf as well as the previous parjdev.pdf which is the advanced stuff with tips.

so I'll say a course + 1 month (max) hands on should get you well on the way to
PX mastery or close (really depends on the person were talking about this is my estimate for you upon my impresion from your posts here)

(now I need to do this without the official course :( )

Good Luck :) ,
Roy R.
Time is money but when you don't have money time is all you can afford.

Search before posting:)

Join the DataStagers team effort at:
http://www.worldcommunitygrid.org
Image
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

Roy,

thanks for the answer! I am going to get lucky and will be in a PX course on Wednesday. I suppose part of my confusion lies in the naming convention - a "Extender" is something added onto the product, which led me to believe that I would just have to learn some new things and apply them to what I knew about DataStage. The reality would seem to be that the PX is a completely separate product with it's own distinc learning curve.
roy
Participant
Posts: 2598
Joined: Wed Jul 30, 2003 2:05 am
Location: Israel

Post by roy »

well since the whole engine is different and there are lots of new stages for doing things plus the fact you need to keep in mind your seemingly sequential flow is actually parallel under the hood, you find yourself designing the configuration file for resources available per node (usually not to many times I hope) and planing partitioning methods thru your job flow and determining which stage to use for the job to run with better performance and such, kinda keeps you on your toes ;).

green with envy about you taking the course :wink: ,
Roy R.
Time is money but when you don't have money time is all you can afford.

Search before posting:)

Join the DataStagers team effort at:
http://www.worldcommunitygrid.org
Image
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

Roy - that was a pretty lengthy sentence there... I think my class Sensei is going to be French so I am really hoping that I get some good documentation. So why are the first 2 days of class "classical" DS and the last 3 PX? That doesn't make any sense.
roy
Participant
Posts: 2598
Joined: Wed Jul 30, 2003 2:05 am
Location: Israel

Post by roy »

Well,
I bet that manager/director and intro are clasical plus part of the administrator oh and paralelism methods of pipe line and partitioning.
also sequence jobs might be there.
the last 3 are probably the actual PX

Only guessing here ;)
Roy R.
Time is money but when you don't have money time is all you can afford.

Search before posting:)

Join the DataStagers team effort at:
http://www.worldcommunitygrid.org
Image
T42
Participant
Posts: 499
Joined: Thu Nov 11, 2004 6:45 pm

Post by T42 »

Expert Server guy -> Expert PX guy?

I would say a week to get the GUI stuff down pat. About a couple months to get Orchestrate down pat. A couple more to get Custom stages API building down pat.

As long as you remember this: Find an alternative solution before using the Transformer -- you'll do just fine.
roy
Participant
Posts: 2598
Joined: Wed Jul 30, 2003 2:05 am
Location: Israel

Post by roy »

I ment that the estimate I gave will get him at least close to mastering PX In my opinion, othets might only get productive with the same effort.
I never said server expert = px expert, I ment server expert probably needs a relatively short learning curve, since parallel methodology is known and you only need to modify the aproach of hndling it, and learning new stages (even as many as there are in px) should not take to long.

again it is IMHO,
Roy R.
Time is money but when you don't have money time is all you can afford.

Search before posting:)

Join the DataStagers team effort at:
http://www.worldcommunitygrid.org
Image
Post Reply