HANA Transport Mechanism

[Former Member]

Hi All,

I'm new to SAP HANA and was wondering how the transport mechanism works for Views Created in HANA Studio.

I was just reading a Document which talked about CTS+  ( I'm SAP ABAP Developer - I know CTS )- But I Read CTS+ is used for Non ABAP Objects -Can Some Explain what are Non ABAP Objects .

Also as per the document - First I have to create the Views In Packages which are attached to Delivery Unit - Then Download the Delivery Unit to your local PC and then go to STMS and Create a Transport and then attach the Delivery Unit File from Your Local PC and then Release the transport...

Its almost similar to what i do for SAP ABAP objects but the it asks me instantly when i edit or create a new program - I don't have to download and then attach to transport. Do we have any similar mechanism in SAP HANA or its just the downlaod attach mechanism

Also what about the tables that i created in HANA Studio and i believe the tables are not attached to Delivery Unit - How do you move your tables from DEV to QA to PRD

Thanks

[Former Member]

Thanks Tarun. In our system developers doesn't have access on _SYS_REPO schema. I am looking for some UI as part of change manager or System window where developers can check the Change_id for any object.

[Former Member]

Hello P B,

Unless you have SELECT  ON schema _SYS_REPO inside object privileges for your user , you cannot see repository related details.

Change Manager view or _SYS_REPO.CHANGE_ENTRIES  gives you objects created under a change id , in order to see the subsequent changes made by different user under a change id , you can do select on table SYS_REPO.CHANGE_CONTRIBUTORS.

-Siva

[Former Member]

Hi Everyone,

So glad to find this discussion. Now that we have developed our 'Content', we are on SP06 and would need help to know for any additional features/methods for control and traceability that we may have for a robust transport strategy (apart from CTS+).

A similar discussion - http://scn.sap.com/thread/3245339

I have exactly the same gap for control which you raised. In BW, when we transport a particular object from Dev to QA, the next time you try to edit that object, it prompts for a request. This is a Control measure which is essential. Do we have this in HANA yet?

Another thought that runs to my mind is HANA being a part ad-hoc analysis kind of an enviornment,

if I do not allow any changes in my Production HANA environment, is it logical to go back to Dev do the changes and then transport it through multiple system landscape until I can experience the change? The change I am talking of here is as small as having one more field in my Analytic View and analyzing the information for business support.

Let me know your thoughts please.

Regards,

kG

[former_member182302]

Hey Kunal,

Have a look on Change manager feature in SP7 ( I still didnt get a chance to try this out still )

Development Perspective: Change Manager (SPS7) | SAP HANA

Regards,

Krishna Tangudu

[Former Member]

Thanks Krishna, addition of this tries to bring things inline to the proven ABAP/BW transport request process.

1. As per the video, this will show only the open change requests of the Owner.So that any further developments can be included in the same request or a new request can be created.

2. I did not quite understand the including of new owners concept - this can mean, that other owners (if added) may include theri developments to this request. But does that also mean they can change the objects of the primary owner (may be delete an existing object in the request)?

3. The point raised by Ravindra & me, is how can it track the changes of an object which is already transported to the Quality system for eg. and it should pop up for a request and capture all changes in this new request. Just like BW.

Regards,

kG

[Former Member]

Hi,

With Hana Life Cycle Manager, you can:

1. Capture a repository object in a change id.

2. You can have multiple objects captured in the same change id.

3. You can have multiple developers contributing to the same change id. This will allow one user to edit the object created by some other user (of course only if the user has authorization to edit that object)

4. These change ids get associated to delivery units.

5. While transporting the delivery units you have an option either to select all the change ids or select only some of them.

6. When you move the change id to QA from development and try to edit and activate the already moved objects, you get a popup to associate it to a new change id.

Hope this helps to those who refer to this post going forward.

Regards,

Piyush

[Former Member]

Thanks Piyush for the explanation. Its really helpful.

Is there any way to check the Change Id under which object is currently tracked ? System ask for Change_Id only during first time activation of object. After this object will be tracked under same Id until Change_Id is marked as released. During subsequent changes by different do they have any option check Change_Id under which object is getting tracked ?

[Former Member]

You can find the objects and change IDs from the table _SYS_REPO.CHANGE_ENTRIES .

[Former Member]

Hi PB,

follow these steps if you can not fire a query on Sys_repo table:

1. Go to Change Manager

2. Select the correct server details from drop down

3. Select the user id who activated the object from the drop down. Alternatively you can also select blank user id(the first in the list). By default your own user id will be selected.

4. Select Status as Open and select Search or Go button

5. At the bottom of the screen it will list down all the open change ids. Expand the button beside each change id One by one.

6. Once you expand a change id, you will see objects and contributors. Expand object and see if your object is present. If not then go to the next change id and do the same.

this is a tidious way to check but helpful way if you can not query sys_repo.

[Former Member]

Any Suggestions

[former_member184768]

Hi Venki,

Currently when you modify any object, HANA doesn't track that the object is modified and should be locked for transport. Also there may not be a possibility to transport an individual object. When you collect the package for transport, all the object are collected for transports.

Hence you should consider object collections in the packages and assign them respectively.

My personal preference for table creation is through scripts. These scripts can be executed on different environments. Also some of the ETL tools also create tables and these can be configured for different environments accordingly.

Regards,

Ravi

[thomas_jung]

>Currently when you modify any object, HANA doesn't track that the object is modified and should be locked for transport.

Actually this isn't exactly true.  Changes are tracked.  When you export the delivery unit there is an option called Filter by time.  You then put in the start and end date/times and only those objects created or modified during that window will be exported into the delivery unit.

As far as tables, we recommend that you create those in the repository via the hdbtable file type. The hdbtable will create or alter the catalog table when activated.  You can transport the hdbtable via a delivery unit as well. In fact just about all the catalog objects (schema, tables, views, sequences, etc) have repository representations now in SPS5. 

[former_member184768]

Thanks Thomas for the clarification.

I was commenting from BW perspective where once you edit any object, it is automatically prompted to put on a Transport Request.

I must admit that currently my knowledge on Transport mechanism is not very much as we are still in process of finalizing the mechanism.

Regards,

Ravi

[Former Member]

Hi Ravi and Thomas

I just want to know what type of Mechanism most of you are using

1)Export Delivery Unit from DEV- Import Delivery Unit File to QA -  Import Delivery Unit File to PRD

OR

2) )Export Delivery Unit from DEV - Attach to transport (CTS+ Loose Coupling) Release the Transport Request Number- Import the transport to QA - Import to PRD

Also can anyone tell how tight coupling works

Thanks

[mathanponnucham]

Hello,

You need to manually attach the HANA development objects and transport it across systems, below info is available in page 19 of SAP HANA Technical Operations Manual (http://help.sap.com/hana/hana_tom_en.pdf .

Controlling Change

In integrated system landscapes, it is important to manage all changes in a central system. This is the only way to synchronously execute changes that affect more than one production system. For example, if you use SAP HANA application-specific accelerators (e.g. CO-PA) in combination with your SAP ERP backend, you might need to synchronize changes of the data models on SAP HANA with changes for accessing the changed models from the transactions on SAP ERP. SAP Solution Manager provides central transport functions for the entire system landscape as well as additional support for quality management.

The process for transporting changes from a SAP HANA development system to a quality assurance and further on to the productive system starts in SAP HANA Studio. There you retrieve content from the HANA source system and export it to enhanced Change and Transport System (CTS+) in SAP Solution Manager. The exported delivery units are attached automatically to a transport request of CTS+. After you have released this transport request, CTS+ triggers the automatic activation of the change in the SAP HANA repository on the target system (quality assurance or production).

Besides supporting the transport and activation of new or changed data models, views, stored procedures, or analytic privileges, SAP Solution Manager offers additional possibilities to improve the quality of changes. You can use change analysis and reporting for getting information about the current status and history of changes. Configuration validation checks whether the systems of your system landscape (e.g., development, quality assurance, and production) are configured consistently and according to the requirements. Quality Gate Management provides an additional quality inspection for projects and ensures changes are transported correctly and synchronously to the production systems. And Change Request Management within SAP Solution Manager controls the entire change execution with detailed process and workflow control. Out of these options, you can select those for controlling changes on SAP HANA that fit best to the quality strategy of your organization.

Rgds,

Mat.

[former_member205400]

Thomad Jung,

RE: saving the table definition to include in a delivery unit !!!!

I'm on ver 1.0.60 and the only way I can see including the table definition in a delivery unit is to create a procedure in the content package. Is that what you are talking about?

On more thing ..... the ability to pull objects "BY TIME" and putting them in a transport is far from satisfactory. We have the need to put objects in a package and then only pull a portion of those to the transport, e.g. I have 6 objects, 2 in development, 4 ready for transport .... I have to go in to the 4 and edit them and pretend to change something so I can specify a "BY TIME" to put them in a transport. I think someone didn't think this through. Also, If i have a procdeure that creates, alters, etc a table definition I can't be putting that in a transport where I'm modifying a view !!!

Mike

[thomas_jung]

>I'm on ver 1.0.60 and the only way I can see including the table definition in a delivery unit is to create a procedure in the content package. Is that what you are talking about?

Not at all.  In SP5 we introduced the HDBTABLE development artifact and in SP6 we have the additional artifact called HDBDD.  Both can create tables via the Repository. Refer to section 4.2 and 4.7 of the HANA Developers' Guide for more details.

>On more thing ..... the ability to pull objects "BY TIME" and putting them in a transport is far from satisfactory.

We agree. That is why in SP7 we are adding change request objects so that you can manage on an object level which groups of objects will be transported.  However, until then you only have the BY TIME option.

>Also, If i have a procdeure that creates, alters, etc a table definition I can't be putting that in a transport where I'm modifying a view !!!

This is why you should create these objects in the repository instead.  When transporting a delivery unit, the dependencies are analysed and processed in the correct order.  Therefore you can transport table definition changes/creations in the same DU as a view which uses that table.

[Former Member]

Hi Thomas,

CTS + can only be used for Delivery units ( whole package)

Using delivery units and CTS + we can only transport the whole package.

We have more than 30 analytical views in HANA package but we have been making changes to several analytical views and want to transport only one analytical view so the BY TIME option is not useful.

If i want to transport just 1 analytical view then i have to use DEVELOPER MODE to export and import as the DEVELOPER MODE  is not integrated with CTS +.

Do you have any suggestions... how to move just one   analytical view using DELIVERY UNIT as we have CTS + configured.

Regards

Satyam

[thomas_jung]

The new Change request functionality in SP7 is not yet integrated into CTS+.  It would solve your problem, but can only be used with HALM routes. We plan to offer integration with CTS+ in next support package. So until then you will have to use BY TIME. You can go into the single view you want to transport, make a very minor change, and then use the time window of that change to force export just this one object.

[Former Member]

Hi Tom

    I have created a couple of SQL script files under the HANA project folder in the development perspective. We created a delivery unit and exported the package related to the project and then imported the delivery unit into a QA system. Where can I find my SQL script files? Should I  expect to see the repository created in the QA system and should be able to view that in the Repository view in the development perspective ?

   I will highly appreciate your help in this regard.

Thanks

Saleem

[thomas_jung]

> Where can I find my SQL script files?

In the exact same place they were in the development system.  The DU imports the packages and their content and activates the content (creates the catalog objects).

>Should I  expect to see the repository created in the QA system

Repository created?  Every system has a repository by default.  The import process doesn't have to create a repository.

>in the Repository view in the development perspective ?

If you create a connect to the repository on that system and if you have access rights to view the repository - then yes.

[Former Member]

HI Tom

   Thanks for responding. May be I should have phrased the question better. I have created SQL scripts under a project in the development perspective which I have linked to a package which is linked to a DU. When I import this DU in the QA system, is this project automatically created in the QA system and can I expect to see the SQL script files under this project in the QA system ?

   Also when the DU is imported, will the system just activate the SQL script files or will it run them as well.

Thanks

Saleem