A few months ago an SAP HANA pilot for the QA environment was setup for a couple of scenarios in the customer where I am currently working, I was not involved at that time in the project and almost all the work was done by SAP consultans. Tomorrow we will be starting a new phase in order to move it to production, the technical part started a little earlier to meet the requirements though and since I underwent the SAP HANA Installation and administration training I got involved in the project. This is the list of what we were requested to have accomplished:
– ERP & SLT server with DMIS 2011 SP3
– Implement SAP Notes specified in Note 1759156
– Create an RFC user for the connection with proper authorizations
– Create a tablespace for logging tables
– Create an index for the involved tables with over 10M records in the key field that has the most number of unique values
– Make sure ports used by HANA are not blocked
– Setup SAP remote support connection
– At least 1 Gbps connection between SLT and ERP servers and between SLT and HANA servers
– SBOBJ Platform 4.0 SP 2 Fix 5 or higher.
The architecture consists of an SAP ERP source system, an SLT Replication server, the SAP HANA system and SBOBJ.
I had some e-mails exchanged with SAP regarding the indices because in the training we were told that when we want to change the reading type to 1 (Access Plan Calculation) based on the first key field then an index is not required, the SAP consultant replied that based in their experience the index still is required to speed the load, for a customer they had running a load for 3 days with only 150,000 records replicated, once an index was created the whole table with 700,000,000 records was loaded in 3 days.
While we are finishing to meet the requirements the replication of some tables was started again by the SAP consultant in the QA environment so the functional consultants could start working and tomorrow I will do it to for some other tables for what will be my first steps in this new adventure…