Transport Procedure???

[Former Member]

Dear Friends,

I have a confusion in the transport procedure in BW. In BWDEV BC, when i activate InfoProviders (say in FI-AR),i get a "assign a request screen" and i assigned a new request to it. Also when i make some changes in the BW DEV system (like deleting a source system assignment) i get an assign request screen and i assign it to a new request again. Now when i am done with all the development in the BWDEV system, what should be approach for transport to BWQ? Do i need to transport the requests that i created earlier or do i just go to transport connection and start activating at the infoprovider level with dataflow before? Please advise.

Thanks

Raj

[udayabhanupattabhiram_cha · Accepted solution]

Hi Raj:

Especially in BW, you can NEVER transport all objects in a single step in a small size project.

The reason is:

Almost all the time, I have some development that depends on BW Objects.

E.x. I create a Cube. I want to extract data from this Cube to another Cube for further transformation or consilidation or Backup.

In this case, I first have to transport my Source Cube. Create a Generate Export Data Source in QAS. Then, replicate data sources for BW QAS Soruce System. Make sure I can see this replicated Data Source in QAS. Only then I can transport new upate rules for second cube.

Also, don't ever transport Queries with Cubes, not a good practice.

We can actually write a small book in transports in BW as they are very diferent than R/3.

NOTE: When some old school R/3 people tell you to release transports without actually transporting them, say NO to that practice. Because, especailly with transported Process Chains, you will need a create a TEMP requent whenever you try to activate & schedule it. These will add unnecessary release requests.

Keep learning about transports because your transport strategy says a lot about your developemnt skills.

Good luck

Ram Chamarthy

[Former Member]

Ram, Thanks a lot for the input.

Raj

[Former Member]

Hi Ram,

As i understand from you:

1. you assign everything to requests during development.

2. Delete those requests (Unlock from SE03 and delete in SE01).

I am confused with part 1 and 2 . In the transport connection do you collect objects individually (like infoobjects, infoprovider) with option "with necessary objects only" or

do you collect part 1 at once (with option???) and part 2 at once? Please clarify.

Thanks

raj

[Former Member]

Hi Guys,

Thanks to all of you for your prompt answers. AHP, I installed the infoproviders (in BC)with “in dataflow before” and I got the screen to assign it to a request. Can you please clarify where these transport settings are in BW ? Please explain what you meant by “since system ask for request, or your system has setting 'record request'-in scc4”. Also if I choose to transport everything through the transport collection, should I choose the InfoProvider with “in dataflow before” or “only necessary objects”. Please suggest?

Thanks again for the help.

Raj

[udayabhanupattabhiram_cha]

Hi Raj:

I was very lucky with my BW transports becasue I follow a different system. I very rarely get failed transports. but it takes some tims and thinking to get used to it.

I will explain you my process in brief.

1) While development, Eventhough i assign a transport request, I don't release it. BW is different from R/3. Till i am ready for transports, i leave all the requets in couple of temp requests.

When I am ready for transports, I go to SE03. Unlock all of them (Start from Inside and move up to the main Request). If someone has ownership along with you, you need to speak with them and take ownership.

As soon as you unlock, delete the requests in SE01. This is very impotant because Unlocking without deleting will create inconsistancies.

Then I start the collection in Workbench.

From here, there are many methods to kill the bird.

I follow this:

Part-1

InfoObjects.

InfoProviders.

infoSources.

Update Rules.

Generated Export Datasources

Replicate Source Systems in QAS

Part-2

InfoSources attached to the Replicated Data Sources.

Queries, etc.

You will get used to it and will have your own flavor soon.

Happy transporting.

Ram Chamarthy

[edwin_harpino]

hi Raj,

how do you activate the infoprovider ?

with 'in dataflow before' ? and you use option 'install and transport' ? (since system ask for request, or your system has setting 'record request'-in scc4)

you can check the request, expand, see if there are infosource etc. if all objects are there, you can release these 2 requests (se09/se10),

and go to target system, and STMS to import request.

if not so, you can go to transport connection and collect with 'only neccessary objects' and then 'in dataflow before'.

remember to maintain the source system mapping

and replicate datasource from r/3 in bw target (qa/prod)

hope this helps.

[Former Member]

Raj

Basically if you want to see the changes you done in Development box in QA and Production system you have to transport all the requests which you create during development stage. in your case you have to transport both the requests with overwrite capabilities functionality that means it over writes all the changes previously send to QA and production systems.

Hope this helps

Thnaks

sat

[former_member184494]

Raj,

if you go to STMS or SE10 , you will see all your requests and the object list against them as well.

My suggestion is that , if you want to move them in toto , merge the requests and then transport the request in one shot , otherwise , you will have 'Child' requests below a main request , release all the 'Child' requests and then the main one and then import the main one to the next system.

Hope this is what you were looking for...

Arun

Assigning points is a way of saying thank you in SDN