ASUG Webcast: Migrating to BW on HANA: Q&A Session
Today we had an interactive Q&A webcast with SAP’s Storm Archer and Brian Woods. Special thanks to ASUG Volunteer Allison Levine for moderating.
ASUG Webcast: Migrating to BW on HANA: Interactive Q&A Session
Date: Wednesday, April 08, 2015
Q: if we have already archived on older technology (ADK and Open Text), how do we get the data back after upgrade from 7.0 to 7.4 BoH
A: Data that has been archived does not need to be restored unless there is some business reason, in which case if it is restored prior to the upgrade and migration, it will be available in the new 7.40 on HANA system. This can also be done in the case where you want to upgrade but are not yet ready to migrate to HANA. Note number 1652907 (next slide)
Q: We are planning to upgrade in next 2 months and stuck on archiving question. SO WHAT’S THE QUESTION?
A: We often recommend a NLS project along with a general “clean up” project prior to the Upgrade and Migration. Cleaning up any nonproductive objects, and moving static “Cold” data to NLS will reduce the RAM requirements for the HANA Appliance, and reduce the time for the upgrade and migration.
Q: Can we migrate multiprovider to composite provider without loosing queries and query elements?
A: No, not currently.
Q: As in the current stage, can the eclipse based UI do everything what can be done from the conventional modeling experience in app layer ? Anything which has been discontinued ?
A: The Eclipse based modeling tools do not provide All of the functionality of the traditional UI at this time. All new modeling objects and capabilities are being addressed by these tools, but we have not made an effort to migrate all existing functionality to Eclipse. Nothing has been discontinued at this point, and there are no immediate plans to do so.
Q: If we upgrade to BW 7.4 SP8 (or later), but not yet move to HANA, what benefits will we receive — or not receive?
A: Slide 38 in the Roadmap Session shows the new functionality added to SP8-SP9 and indicates which features are for HANA only, and which release the functionality was made available in. This is only the newest functionality, there are also slides like this for the SP05 features and functions which I can share during the Q&A session.
Q: In the current Infocube/DSO world, the size of a DSO is about 4 times the size of a cube containing the same data (due to the sid’s vs the native characteristics being stored). For ADSO’s which are cube like is the size more like a DSO or a cube?
A: There are a number of considerations that go into the size. Depending on the type of data will affect the overall compression. The overall goal is to compress data where we can and improve overall performance.
Q: If you create Bex query in BWMT can you change in Current Query designer?
A: The BWMT tools are used to create BW objects and elements stored in the BW system. When you create the Bex query, it is stored in the BW Schema and should be accessible for update inside the SAP GUI.
Q: How these HANA / BW modeling are affected when it goes onto S/4 on HANA?
A: All BW Info sources should continue to function. It is not yet clear if the extractors will be modified, or if they will be replaced with Core Data Services (CDS), but while the mechanism may change somewhat, the result will be the same. As many of you know, this is not yet determined since the Suite/S4 HANA team will have to make the necessary modifications to ensure consistent data and harmonized data models between S4 and traditional Suite.
Q: Does the cost of NLS comparable to HADOOP?
A: NLS sits on top of a data persistence technology. We believe that SAP IQ is the best technology for this, and accordingly our solution is based on this. We do have Partner solutions that are based on HADOOP, and I Would be happy to have a call to discuss the fully loaded costs of each if you like, but the summary is “Use the right persistence technology for the job”
Q: where do we define which data goes to warm storage? is there any T Code?
A: For BW objects, the only objects that are available for this Warm Storage (based on Extended Storage/ Dynamic Tiering) are PSA and WO/DSO (Corporate Memory). The option is set when the InfoProvider is Activated. It is a check box on the modeling screen.
Q: Can you discuss how the ODP removes the need for the PSA? Since the ODP is on the source system – what version/release is required in your ECC system to support the ODP?
A: Not stopping at PSA layer to do transformation; PSA is a staging area not a corporate memory; ODP/ODQ structure can reside on source system or target or separate
Q: Where can we find a list of NLS restrictions besides cannot being updated
A: There is an FAQ for SAP IQ based NLS http://scn.sap.com/docs/DOC-54468
Q: Does SAP BW on HANA 7.4 support 3rd Party NLS Storage like Informatica NL Solution and will it have all the same capabilities that
will be available to SAP IQ?
A: Yes, we still support our partner solutions for NLS, it is up to the partner to decide if they want to be certified for any given release of
Q: Note 2063449:Push down of BW OLAP functionalities to SAP HANA referenced in the presentation is not released to customers?
A: Note is released and available. Details of OLAP functionality and availability is listed some are available and other are planned or
Q: What is required to transition a mature SAP BW installation on Oracle to SAP BW 7.4 HANA?
A: The upgrade and migration process is described in great detail (http://scn.sap.com/docs/DOC-46824)
and you will have a choice to either migrate the existing models as is, or conduct a project to rearchitect
the solution based on the current and planned BW on HANA capabilities (remove InfoCubes, reduce
layers, leverage new CompositeProvider, Open ODS View), and more generally the logical EDW concept
and federation based on SDA.
Q: Once the data has been archived in HANA, how does the memory get freed up? In the past with
traditional databases, it is done through database reorg, how is it handled in HANA
A: To actually purge the data marked deleted (like after a NLS run), a delta-merge completes the process
in removing the data from the HANA Data log files.
Q: Looks like SAP Note 2063449 not released, perhaps being updated
A: See above
Q: We are currently using the 3.x archiving method to archive our non cumulative Infocubes.
Will we run into issues with the archived non cumulative infocubes when we migrate to hana
A: Expanded capability to handle non cumulative cubes. Should not be a problem with new functionality.
Q: What is the Utility to convert from Multiprovider to Composite provider and can I download somewhere on marketplace
A: No, Not currently. We have plans, but no ETA at this point. (updated 5/27/2015)
Q: When we are talking about Role based roll out of GUI, is this SAP Fiori ?
A: Fiori is not a development GUI
Q: Do we have any NLS Project case studies? Where can I get it .
A: The FAQ link is in the slides for NLS. There should be some stories located there. http://scn.sap.com/docs/DOC-54468
Q: What would be the benefits of utilizing dynamic data tiering and NLS vs. using something like a Hapoop for a cold data storage?
A: Dynamic tiering is a capability of HANA, not Hadoop. This dynamic tiering is using disk based tables in HANA. That is the same technology used in Sybase IQ. You can use it but you will need to write the application do make this happen.
Q: does the SAP NLS solution have any advantages over the partner solutions?
A: Yes. SAP can take advantage of new featuers in HANA and BW much more quickly than partners. For example, able to use proxy tables to speed up processing of NLS queries by a factor of 3 times.
Q: Is there a longterm advantage of using SAP IQ vs a 3rd Party NLS solution?
A: Having one vendor for all of your system is an advantage. Also see above response on SAP NLS advantages.
Q: how will SAP be able to help if we want to migrate data from an existing 3rd Party NLS system to SAP IQ?
A: All the data will have to come back into BW and be re-archived.
Q: Do you see any real viable use case for using smart data access with Hadoop and BWoH?
A: Yes. SAP is rated as a leader in advanced analytics (predictive modeling). For the predictive modeling you don’t always know what the data will look like.
Q: IS the preferred NLS solution is to use Dynamic tiering instead of Sybase IQ?
A: No. SAP NLS solution still runs on a standalone IQ system.
Q: Whats the advantage of NLS solution like Sybase IQ if we implement Dynamic tiering? I thought we culd use Dynamic tiering for Warm data and use IQ for archiving(COLD Data)
A: NLS is not the persistence technology. You still need something that maintains the data and insures query integrity.
Q: So NLS is integrated for historic reporting.?.so it holds mainly the reporting layer data..and not the PSA data..
A: Yes correct
Meet Storm and Brian next month at ASUG Annual Conference – see Find ASUG BW HANA Content and more – at ASUG Annual Conference
Join us May 4th
BI pre-conference session at ASUG Annual conference
Monday, May 4. (extra registration fees apply).
Featuring Hands-on SAP BusinessObjects BI 4.1 w/ SAP NetWeaver BW Powered by SAP HANA – Deep Dive
REGISTER Limited to 30 people.
Check out the ASUG Education Sessions at Conference: ASUG BI Schedule 2015.xlsx
Several upcoming ASUG BI Webcasts includes BI, BW, HANA, EIM & more – register