DEV/TEST/PROD Scenario for Extractors

dataservices sap-erp-bw
1

Hi

Just wanted to throw this and get an idea on this…

We came to the level of using Extractors in DS14.

Now really managing and trouble-shooting is a kinda question that runs in my mind.

Delta Queues are really kind of Message Queues in cases where a Record is snatched from the Queue, it is all done and the record will disappear.

So if something isn’t worked the way expected in PROD, how to simulate things in DEV, as the data is now gone? I just want to brainstorm or get a consensus on that?!!!

Also, how much burden does it creates on having queues for TEST & DEV in a PROD Box and at least a DEV queues in the TEST Box.

How much burden on manual maintenance? I really want to get an idea as we are travelling to a PRODUCTION World, really seeing issues on initial data and having hard time troubleshooting…

ganeshxp :us: (BOB member since 2008-07-17)

2

Can someone give an idea to this please?

I generally assume that the system is capable of doing things provided the resources are given :roll_eyes:

But the guy who owns the Extractors for Data Services says that this is a too much burden to the SAP System on having multiple queues on the system?

How much burden can it create for SAP Tables that holds data in few million and expected to give delta data in some 10’s of thousands everyday? Really this is confusing my mind.

ganeshxp :us: (BOB member since 2008-07-17)

3

That’s a tough one as it can be everything.

An Extractor is an ABAP program usually and what that program does is up to the programmer of it. Today we came across one special CRM extractor which runs for 2 hours to load 6000 rows into BW.

But actually, your problem as such is not related to Extractors alone. How do you debug a table read when the prod system has a very specific combination of values and your dev does not?

I personally would keep things clear, prod for prod, dev for dev.

Oh, btw, once a change is read, it does not get removed from the queue right away, there is a SAP batch process that does that. In order to re-read an already processed delta, you have to set the extractor timestamp field in the Extractor Reader to your delta start date and when it contains an old value, the old delta will be read again, assuming it exists still.

Werner Daehn :de: (BOB member since 2004-12-17)

4

Hmmm then I would say wrong preaching!!! :hb:

So if those wouldn’t be gone, then can i read against the same queue if the timings are setup properly.

I mean, if I get a data issue in PROD, I can point my DEV/TEST job to point to the same PROD queue and see what the problem is? Will that work?

Do you mean setting in the Data Services or should I configure something on the SAP System?

ganeshxp :us: (BOB member since 2008-07-17)