Automizing parameter values with a concurrent architecture

ds_is_fun · Post by **ds_is_fun** » Wed Sep 26, 2007 11:10 am

We currently have an architecture where we will be running batch loads based on a RUNDATE on a daily basis.
There are three stages that the data batch needs to pass through.
Stage 1. Source - Standardization
Stage 2. Standardization - Cleansing
Stage 3. Cleansing - Target
In the first stage we extract based on a parameter RUNDATE.

Going forward from Stage 1-2, Stage 2-3 Im currently placing the RUNDATE, Batch_LoadID in a hash file for all the transformation required.
If another job is started at the sametime then it is possible that it would incorrectly refer to that hash file.
What is a good method to prevent this non-concurrent method in a DS architecture design? where no third party tool is used for scheduling and only Job sequence, DS designer is relied upon.
Thanks

ds_is_fun · Post by **ds_is_fun** » Wed Sep 26, 2007 2:19 pm

pl. answer

ray.wurlod · Post by **ray.wurlod** » Wed Sep 26, 2007 4:16 pm

Learn patience.

You posted at 3am and 6am my time.

Penalty for impatience: 24 hours before I read the post.

If you require urgent support sign up for premium service with your support provider and learn the true cost of "urgent".

ds_is_fun · Post by **ds_is_fun** » Wed Sep 26, 2007 9:46 pm

There was no urgency seeked as remarked. There was tremendous posts today and din't want the post to be lost from "Recent Topics". May be that was crazy but motivation was diligent.
Would appreciate any suggestion on the architecture. No rush! Thanks!