DSXchange

hi,

Presently iam running the production jobs. which contains 130 jobs,
these are all divided into 8 sequences.All to these sequences are add into one main sequence.
Previously the total elapsed time was around 9 hrs,
but presently it was 13.5 hrs. But there is no changes in the jobs. how can i improve the performance of the sequence.

Why has the total elapsed time increased by 50%? Are you processing a greater volume of data? How hard are you working your machine? Are there "running in parallel" or "time shifting" opportunities to be had to even out the load on the machine? These are just some ideas to get you started.

Yes, greater volume of data but not that much.
At the time of development they are using only joins and at Query level.
they are not using any other. All the logic was inserted in query. Not using any stages only transformer stage.

4friends wrote:Yes, greater volume of data but not that much.
At the time of development they are using only joins and at Query level.
they are not using any other. All the logic was inserted in query. Not using any stages only transformer stage.

Havn't got all the answers

Greater volume of data will increse the load time.

Some Questions (1st one inherits Ray's Question):

1. Are you running jobs in parellel?

2. Is there any thing else consuming the resources while you are running your jobs?

Few Suggestions:

1. Try to call the maximum number of jobs at a time if the resources and dependecies permit.

2. Tuning your queries may reduce the load time.

Regards,

You have some work to do. Where did this increase come from, any jobs in particular? Identify the culprits, then you'll know where to concentrate your efforts.