Recording is here BI: Data Tiering Optimization with SAP BW/4HANA

Slides are here

Hosted by Enterprise Data Warehouse SIG

Data Tiering Optimization in BW/4 HANA offers the possibility to automate the distribution of hot, warm and cold data to the appropriate storage media.The presentation will provide an overview on the configuration steps and an update on the planned roadmap for DTO.


Josh Djupstrom, SAP


Don't miss our upcoming webcast TODAY BI: SAP Analytics Cloud for Beginners

Also June 4th please consider joining the ASUG BI Community for Jump Start ASUG Annual Conference 2018 SAPPHIRENOW with SAP Analytics Cloud/ Lumira Designer Session


Question & Answer


can the warm store be the disk on the HANA Appliance itselfThe extension node effectively is disk storage for warm data. The extension node is part of scale out, but has a lesser hardware requirement than the hot nodes. Then there is some configuration for the extension node that specifies it as warm storage.
Can the disk on the same HANA Appliance be used as warm storage in a BW4HANA scenario?Yes. The extension node uses the appliance disk store.
any license impact for extension nodesThere is likely some charge for the extension nodes, but the node is scaled down in CPU/RAM compared to HOT nodes. Any licensing costs/configurations have to be discussed with the Account Exec.
Is the data to RAM 50 to 50 only applicable to SAP BW/4HANA?I presume you are referring to the hot nodes. This is a standard for any BW/4HANA or BW Powered by HANA.
What is the ratio for an Extension Node?  80/20?There is not; something you can play with
How much of this information applies to BW 7.5?This still applies. There is a slide coming up that will show the 7.5 versus BW/4HANA details.
Is the time characteristic require to be a key field?Yes it does.
hi, this is Sudeepti, WARM is on extension nodes, extension nodes are on? not clear on scale out nodes, something faster than disc?The extension node is another node on the HANA appliance. That node does not have the same CPU/RAM requirements the HOT nodes have. There is configuration in the HANA appliance that specifies that node as an extension node.
the classification of data is basis only dates? if no, what are the other options to define data temperature?There are more slides coming up. Please re-ask if the following slides don't answer the question.
thank youYou're welcome
Can we expect DTO for SOH systems?DTO is for BW systems. I don't know if SOH or S4 will have a similar function out side of the regular DAP process. That team is the best one to answer this question.
Is it possible to use temperature maintenance to archive data based on Non-key fields (i.e. Data fields) in an ADSO? in the demo CALDAY was a key field on the ADSO so trying to understand the scenario where there is no time reference characteristic as key field.Josh is present now on the best practices. Please re-ask if you don't get your answer.
If data from the aDSO needs to be loaded to another aDSO, what impact does the the different temperature settings have what data is available to be transferred?The access to loading from an aDSO to another is transparent. The only difference typically is the speed with which older data is accessed.
Does the temperature needs to be time based or can it be based on another dimension like product line, sales campaign etc?Does the current slide answer your question or do you need further answer?
Are there tools Or o analyze & suggest partitioning/Temp settings on objects, or if you have poor performance due to bad settings? Or is that now in the early watch reports?Hi Jay - The future direction slide is coming up. There are plans, but not available yet.
from where can we get access to the videos of the demos?The videos should be available after the webinar via a link to our EDW Community Page.
Hi, what is the use of parallelization setting and the difference between adjust, repair and "adjust and repair" settings in the Job Execution screen of DTO? Thank you.ties into partitioning - Josh to look into - capabilities now
So our COA DSO has doc # not date in key. Are you saying we could add an immutable date would allow us to partion (for NLS, etc.) ?as in slides, the date needs to be in key; immutable character would work