Skip to Content
Technical Articles
Author's profile photo Takao Haruki

Transferring data from On-Premise SAP HANA to SAP HANA Cloud using SDI

(Japanese version)


In this blog, I will transfer data from On-Premise SAP HANA to SAP HANA Cloud using SDI (Smart Data Integration).

SAP HANA Cloud includes SDI which is an ETL tool, this means that you can easily integrate various kinds of source systems into SAP HANA Cloud without installing ETL application. Moreover, you can develop ETL jobs as same as developing tables and views, this makes it easy and quick to develop ETL jobs.

I will transfer 3 tables on On-Premise SAP HANA to SAP HANA Cloud.

Let’s get started. The steps are as follows.

  1. Configure Data Provisioning Agent (DP Agent)
  2. Create Virtual Table
  3. Transfer data


1.  Setup DP Agent

1.1 Install DP Agent

Install DP Agent on On-Premise side so that SAP HANA Cloud can connect to On-Premise SAP HANA. In this blog, I will install DP Agent on a Windows server which is in the same network as On-Premise SAP HANA. For more information on installation procedure, please see SAP Help and “Download DP agent from SAP Service Marketplace as a ZIP file” part of this blog.

1.2 Configure DP Agent

Configure DP Agent as following to SAP Help.

1.2.1 Create a DB user for Agent Admin

From SAP HANA Cockpit of SAP HANA Cloud, create a HANA DB user who has following authorizations.

Application Privilege :

* The Application Privilege does not exist and no need to set it although it is written in the help

1.2.2 Run DP Agent Configuration tool

Run command prompt as administrator and configure environment variables as followings, then execute DP Agent Configuration tool.

setx DPA_INSTANCE "C:\usr\sap\dataprovagent"
cd bin
agentcli.bat --configAgent

1.2.3 Start Agent

Choose option 1 (Start of Stop Agent) and then choose option 1 (Start Agent). DP Agent will start by this.

1.2.4 Configure SAP HANA Connection

Configure to connect from DP Agent to SAP HANA Cloud. First, choose option 6 (SAP HANA Connection) and option 3 (Connect to SAP HANA via JDBC). Next, input as followings.

Use encrypted JDBC connection : true
Use WebSocket to connect : false
Host Name (for Direct SQL) : SAP HANA Cloud InstanceのEndpoint(*)
Port Number (for Direct SQL) : SAP HANA Cloud InstanceのEndpoint(*)
Agent Admin HANA User & Password : Information about the user created in Step1.2.1
Use HTTP Proxy Server : false (for this blog, select based on your environment)
HANA User for Agent Messaging : Any user name & password, true (create new user)

Configuration of connection between DP Agent and SAP HANA Cloud has been completed.

* You can confirm the hostname and port number of SAP HANA Cloud in SAP Cloud Platform Cockpit. Please see Endpoint of SAP HANA Instances.

1.2.5 Register DP Agent

Go back to the top of the Configuration tool and choose option 7 (Agent Registration) and option 1 (Register Agent) to register DP Agent.

Agent Name: any name
Agent Host: IP address of the server where DP Agent is installed

1.2.6 Register HANA Adapter

Register “HANAAdapter” which allows SDI to connect to SAP HANA. Go back to the top of the tool and choose option 8 (Adapter Registration). Then choose option 2 (Register Adapter) and input as follows.

Adapter Name: HanaAdapter

1.2.7 Confirm the configuration

Go to Database Explorer and confirm Agent and Adapter are successfully registered.


2. Create Virtual Table

Configuration of DP Agent to connect between SAP HANA Cloud and SAP HANA On-Premise has been completed. Next, I will create virtual tables on SAP HANA Cloud which access to On-Premise SAP HANA tables.

2.1 Create a Remote Source

At Database Explorer, right-click on “Remote Sources” and choose “Add Remote Source”.

Input Remote Source Name and choose “HanaAdapter” as Adapter Name. Input other necessary information like hostname and port number of On-Premise SAP HANA, user name and  password. Save it when completed.

Open the remote source by clicking the name, schema is displayed. And when you open the schema, you see the tables in the schema. The source schema is “SALES_1”, so you can see the tables (SALES::HDI_MD_PRODUCTS, SALES::HDI_SO_HEADER, SALES::HDI_SO_ITEM).


2.2 Grant privileges to create virtual table

Remote source has been created in the previous step. Next step is creating virtual tables. Before this I need to grant privileges to create virtual table to administrator of the HDI container.

In this blog I will directly grant privileges to the user using SAP HANA Cockpit.

First, logon to SAP HANA Cockpit by the user who created the remote source” and go to “User & Role Management” > “Privilege Assignment”. Then, grant following object privileges to a user “<HDI container>#OO”.

  • DROP


2.3 Create virtual table

At this point, I have created remote source and granted privileged to create virtual table. Next, create virtual tables.

Open SAP Web IDE for full-stack and right-click to create a new virtual table.

Input virtual table name, remote source name, schema name and table name. For Database Name, input “<NULL>”. Save and build it, then virtual table is created. Create all necessary virtual tables.

Go to Database Explorer to check the virtual table. Click “Table” and you can confirm that virtual tables are successfully created (VT_ tables here).

Right-click on the table and select “Open Data” to check the data can be successfully retrieved from On-Premise SAP HANA.

Create other virtual tables and also target tables.


3. Transferring data

Virtual tables which access On-Premise SAP HANA have been created on SAP HANA Cloud. I will create SDI Flowgraph to transfer data to SAP HANA Cloud using the virtual tables.

3.1 Create Flowgraph and execute it

3.1.1 Create a new Flowgraph

Right-click on “src” folder and choose “Flowgraph” to create a new Flowgraph.

Empty Flowgraph is created after inputting name of it.

Click “+” icon and you will see various kinds of operators. You can add operators to the Flowgraph from here.

3.1.2 Configure Data Source

Insert “Data Source” to add source table.

Click the gear icon to configure the data source.

Choose “HANA Object” at “Type”, then select source table. Tables will be listed when you input a part of the table name. Choose the source table (virtual table) and click “Finish”.

The metadata of the table is loaded. Click “Apply” to complete the source table configuration.

3.1.3 Configure target table

Click “+” icon and insert “Data Target” to add target table.

Click gear icon to configure data target.

Choose “HANA Object” at “Type”, then select target table. Tables will be listed when you input a part of the table name. Choose the target table and click “Finish”.

At “Setting” tab, configure write type. In this blog, I will check “Truncate Table” and select “Insert” for “Writer Type”. (This means target table is truncated before data loading.)

Click “Apply” to configure mapping.

3.1.4 Configurate Mapping (Projection)

Click “+” icon and insert “Projection” operator to map between source table and target table. Then connect Data Source and Projection and also connect Projection and Data Target.

Click gear icon of the projection, the mapping is automatically configured by the column name. Change the mapping and add filters if needed.

3.1.5 Save, Build and execute the Flowgraph

Click “Apply” to go back and then save and build the Flowgraph. Finally, click “Execute” to run the Flowgraph.

Confirm the job is successfully executed.

3.1.6 Check the result

Go to Database Explorer and open the data of target table. You can confirm the data is successfully transferred from On-Premise SAP HANA.


3.2 Create and execute (a little complicated) Flowgraph

3.2.1 Configure Join

Next, I will create a little complicated Flowgraph which joins two tables. Add two Data Sources for two tables (SO_HEADER and SO_ITEM), then connect them and “JOIN” operator.

Click gear icon of the JOIN operator and configure join condition at “Criteria” tab. Set SO_HEADER as left table and SO_ITEM as right table, then choose “Left outer join” as Join Type.

Write join condition at Join Condition Editor.

3.2.2 Configure mapping(projection) and data target

Go back by “Apply”, then add “Projection” and “Data Target” operators. In Projection operator you can set filters and configure mappings.

In Data Target operator, choose “Template Table” and input name of the template table at “Object Name”. This means that the Flowgraph creates table when the job is built.

Configure data update method at “Settings” tab.

3.2.3 Save, Build and Execute the Flowgraph

Save and build the Flowgraph, then click “Execute” to run the job.

3.2.4 Check the result

Go to Database Explorer and open the data of target table. You can confirm the data is successfully transferred from On-Premise SAP HANA.



Transferring data from On-Premise SAP HANA to SAP HANA Cloud using SDI has been completed.

As mentioned at the beginning, you can use ETL in SAP HANA Cloud without installing ETL application separately, as long as you setup DP Agent. ETL jobs can be created with the same development environment and development flow as creating table or view, which can significantly improve development efficiency. There is a complete set of operators required for ETL, and it is possible to integrate data flexibly from various systems.


Assigned Tags

      You must be Logged on to comment or reply to a post.
      Author's profile photo Ina Felsheim
      Ina Felsheim

      Thanks for providing these steps. Love the detail, Takao. #SAPHANACloud

      Author's profile photo sudheer keerthipati
      sudheer keerthipati

      Good one??

      Author's profile photo Goutham Nagaraj Naidu
      Goutham Nagaraj Naidu

      Thanks for the blog Takao.

      You have not mentioned anything about cloud connector here , Do we connect to on-premise dp agent without cloud Connector ?

      Best Regards,


      Author's profile photo Takao Haruki
      Takao Haruki
      Blog Post Author

      Hi Goutham,

      We are able to connect to on-premise DP Agent without Cloud Connector.

      According to HANA Cloud L2 document, Cloud Connector is planned to be supported.




      Author's profile photo Goutham Nagaraj Naidu
      Goutham Nagaraj Naidu

      Hi Takoa,

      Thanks for the confirmation , But I wonder how would it connect to on-premise without reverse proxy(cloud connector)  currently.

      Best Regards,



      Author's profile photo Ravi Condamoor
      Ravi Condamoor

      Hi Goutham,

      If the on prem systems are not protected by firewalls, you don't need a Reverse proxy or Cloud Connector. DPAgent can connect directly.

      @Takoa-san, can you confirm that the on-prem system did not have a firewall around it?




      Author's profile photo Vivek Sahu
      Vivek Sahu

      Thanks Takao Haruki  for sharing the steps, it was really helpful.

      However, is there any way we can set below application privilege from Hana Cockpit of SAP Hana Cloud ?
      Application Privilege :

      And from User Management Console of hana cockpit also, I cannot see any such option for "application privilege" to assign it to user.

      It's working fine though on on-premise Hana machine both from Studio & hdsql.

      Author's profile photo Takao Haruki
      Takao Haruki
      Blog Post Author

      Thanks Vivek for the comment.


      I tried again and I also see that the application privilege does not exist.

      (I can see "Application Privilege" tab in Privilege Assignment of HANA Cockpit.)


      I tested without granting it and it worked fine.

      So, can you proceed further without granting the application privilege?

      (I updated this blog)




      Author's profile photo Vivek Sahu
      Vivek Sahu

      Hi Takao,

      Yes, it worked and now I am able to create VIRTUAL table on SAP Hana Cloud.

      As we are not using Web-IDE and from Database Explorer of our Hana Cloud, we can replicate the table successfully by doing following steps :

      - Create Virtual table
      - Create target table
      - Create Remote Subscription
      - Queue Remote Subscription
      - Setup Initial Load
      - Sync remaining changes & Replicate

      Only thing, is there any way to replicate all the tables of any specific SCHEMA together at once?

      Or, we need to follow same steps as mentioned above for all tables ?



      Author's profile photo Takao Haruki
      Takao Haruki
      Blog Post Author

      Hi Vivek,

      As far as I know there is no way to transfer multiple tables in one go.

      I think followings are more efficient than SDI flowgrph (but not in one go).

      1. SDI replication task
      2. Virtual table replica

      Regarding #2, I have created another blog how to use it.




      Author's profile photo David Fernandez Castro
      David Fernandez Castro

      Hi Takao,

      ¿any recommended approach to do a data migration from S/4HANA Cloud to S/4HANA On-Premise?

      Author's profile photo Cesar Felce
      Cesar Felce

      Hi Takao,
      Thanks for the detailed blog, but i'm having trouble connecting from the DP Agent



      I have tried option 3 and 1. But no success.

      [Cannot connect to host [myhostname: myhostname: Name or service not known], -813.].

      Ensure that the specified hostname and port are valid.

      I'm using DBADMIN to connect and create the AGENT User.

      What I am missing?

      Kind regards,



      Author's profile photo Vivek Sahu
      Vivek Sahu

      Hi Cesar Felce ,

      I also faced similar issue in my early days of learning SDI DP agent connectivity with SAP HANA Cloud. Later I noticed that while deploying SAP HANA Cloud instance I have provided limited ip addresses to connect in whitelist entry. So, I changed it to option “Allow all IP addresses” and then it worked. No need to go for Option 3.

      You can cross check the same on your SAP HANA Cloud instance, it seems like similar error message to me.

      Best Regards,
      Vivek Sahu

      Author's profile photo Alejandro Chinchilla Ayapán
      Alejandro Chinchilla Ayapán


      This works for transfer tables and calculation views? or only tables?





      Author's profile photo Nguyen Son
      Nguyen Son

      Hi Takao Haruki,

      I can't see HanaAdapter in dropdownlist? could you help me how to see it or have other choose the same?

      Author's profile photo varun bhargav
      varun bhargav

      Hi Takao Haruki
      Thanks for the blog, really helpful!

      Dose this work inside a HDI container too.


      Thanks and Regards,


      Author's profile photo Jessie Wu
      Jessie Wu

      Hi Takao Haruki

      Thanks for the blog, it really helpful. But I got the issue when I try to set the target table for the  flowgragh. What do you mean create the target table? And what the access grant to the target schema?