The Amazing Analytical Capabilities within SAP Cloud Platform
Over the last month or so I devoted my time on the Analytics with SAP Cloud Platform openSAP course and overall it was a really good experience. Again, like so many other openSAP courses that I have completed, this course offered a great mix of detailed theory and practical elements with some exercises that enriched the learning experience. The main use case used throughout the course was using an SAP HANA DB on SAP Cloud Platform and connecting it via Live Connection to SAP Analytics Cloud to create some great analytic dashboards, which was pretty cool. I was blown away as to how quick this could be achieved. So, given this use case the course not only offered the practical exercises on SAP Analytics Cloud but also covered how to create a HANA DB on SAP Cloud Platform covering the Editor Security and Catalog functions within the SAP HANA Web-Based Development Workbench. I’ve carried this out a number of times before, but I found it beneficial to repeat the process – just embeds my understanding and provides me with more confidence if I need to do this anytime in the future. For that, I was thankful it was covered again ?.
Overall, the course covered a nice range of topics. Main areas covered were of course SAP Cloud Platform, SAP Analytics Cloud, SAP HANA on SCP, as well as the additional analytical offerings such as the SAP Smart Business Service and the Analysis Path Framework Service. The course also had a nice mixture of business and technical elements and like so many other openSAP courses were targeted towards developers, application consultants, technical consultants, and data scientists which is fairly broad.
The course also detailed integration architecture options (I do have a thing for architecture!) for how to connect to live data sources with some information around Smart Data Integration (SDI) and Cloud Platform Integration (CPI). I must admit this is one area that has been confusing to me – especially how to get SDI up and running and for when this option would be used. Previous to the course I thought Smart Data Integration also required the SAP Cloud Connector (as a mandatory component) but after doing this course I now know that this is not the case. While probably not the objective of this course I would have liked an optional exercise of installing the Data Provisioning agent on a backend system and proving that I could set up Smart Data integration. Hopefully a future course could include this although the further reading section provided more than enough resources for those interested to go and check out additional material. This is one area where openSAP courses have really improved. Realistically, the courses cannot cover every single element in detail, but it is good that additional information is provided for those wishing to go further with their learnings. This was the case for not only the integration elements but essentially every topic included additional learnings over and above what was covered. This is seriously great and it is well received by the SAP openSAP community!
From a connectivity point of view a number of options are available between SAP Analytics Cloud and other systems. This diagram is a nice summary of the options. Like with any type of architecture or installations there are different components required depending on the types of scenarios you wish to implement.
Figure:1 SAP Analytics Cloud Connectivity options (from SAP Help site)
The key architecture takeaways for me were the following:
- SAP Analytics Cloud and SAP Cloud Platform are running in the same datacenters. SAP Analytics Cloud essentially runs as a SaaS (Software as a Service) solution on top of SAP Cloud Platform. This was new information to me but good to know.
- To enable Smart Data Integration a small lightweight Data Provisioning Agent is required to be installed in the customer’s on-premise environment (inside firewall). Key to this method of data integration is that the data is not persisted (in the data provisioning agent). This also has limitations to be aware of – check them out here.
- Cloud Platform Integration offers connectivity with on-premise or cloud applications both from SAP and other software vendors. This is the cloud version of SAP Process Orchestration but it can also complement this with the offering of prepackaged integration flows to “Discover, Configure, Manage” to kickstart projects.
The main integration architecture options are covered below – Smart Data Integration and Cloud Platform Integration. I’ve also included a few slides from the course that I thought provided a snapshot of the main elements.
Figure:2 Smart Data Integration architecture (from openSAP course)
I’ve had numerous discussions about Smart Data Integration over the years and the course offered really great insight into this data integration offering. I will summarise my learnings here:
- The target database of Smart Data Integration is the SAP HANA DB. This you can see in the above diagram.
- While there are adapters offered out of the box it does not include all adapters. When an adapter does not exist an open SDK can be used to build custom adapters.
- As described above, data is streamed from target to source and is not persisted (in the data provisioning agent).
- Both batch and real time replication is supported.
- Supports bi-directional data distribution.
- The Data Provisioning server component receives the data.
- While now available for the Cloud Platform SAP HANA service it was originally developed for on-premise databases to integrate different data sources into the SAP HANA DB.
Figure:3 Cloud Platform Integration (from openSAP course)
I’ve had similar discussions about Cloud Platform Integration (CPI) over the years and like Smart Data Integration the course offered really great insight and detail around the capabilities – most of which are detailed in the above slide from the course.
A large part of the course (as it should) also detailed SAP’s Analytic capabilities available in the Cloud. While I did know each of them separately it was good seeing them together and the features of each detailed. The openSAP course summarised these capabilities very well.
Figure:4 SAP Analytic Capabilities in the Cloud (from OpenSAP course)
The core capabilities of SAP Analytics Cloud includes Data Connectivity, Visualisations and Modeling and this was showcased in the course. The main Analytical practical component to the course was to create a Dashboard in SAP Analytics Cloud, this covered some of the navigational aspects as well as setting up of the Live Connection. The SAP Smart Business Service was covered also. I learnt that only 2 models apply in SAP Analytics Cloud – Analytic Models and Planning Models and another key learning was that when the data upload option is used the modeling has to be carried out in SAP Analytics Cloud. This I did not know previously.
The course also covered the other cloud analytic capabilities, namely SAP Leonardo Machine Learning Foundation, the SAP Cloud Platform Portal service and of course SAPUI5 applications that are built with analytical capabilities.
In totality, the analytical offerings are pretty amazing and the openSAP course summarised all of these capabilities brilliantly. I fully recommend this course.
To check this out go to the openSAP website and sign up to do the course or watch the videos in your own time to get up to speed with Analytics on SAP Cloud Platform. Here is the course URL.
How do you get started with Analytics on SAP Cloud Platform you may ask? Well, it is seriously simple – just follow these two steps. NOTE: The SAP Analytics account is a 30 day trial.
- Sign up for an SCP Trial Account -> https://account.hanatrial.ondemand.com/#/home/welcome
- Sign up for SAP Analytics Cloud -> https://www.sapanalytics.cloud/
Post your thoughts on the course or post some of your own dashboards to show us how you are using the Analytical capabilities in the cloud.
Thanks for reading!!!
Glad you enjoyed the openSAP course. Thanks for sharing your feedback.
Yes, there are 3 agents which come into play for such scenarios – Cloud Connector, SAC Agent, DP Agent. Wish they are all consolidated.
When you say “data is never persisted”, I assume you meant for Live connections in SAC and not SDI.
You can try installing SDI DP Agent and replicate data to SAP HANA DB in SAP CP Trial account – https://blogs.sap.com/2016/07/14/step-by-step-setup-sdi-for-your-hcp-trial-account/
Thanks Murali Shanmugham and yes was going to add that blog into the post to provide some more information so thanks for doing that. Yes - with the Live Connection option no data is persisted. I will update the post to make it clearer around the data persistence comment. It is included in the slide in Figure 2 however to clarify it is simply stating that the Data Provisioning Agent itself does not persist the data which is correct. Figure 2 relates to the Data Provisioning Agent architecture specifically so the comment is specific to that.
Thanks for sharing your experience with the Analytics for SCP course! I also did this excellent OpenSAP course and found it extremely helpful for understanding how we can use and integrate all the various analytic service capabilities together in an "intelligent enterprise....without ETLing or "acquiring" data everywhere.
The course covered several SCP and SAC analytic capabilities that I hadn't investigated before, and provided me new context for potential enterprise strategies. I was unaware SCP had built-in and easily enabled analytic services such as Smart Business Service, Analysis Path Framework, along with with integrated Portal & Fiori capabilities...What I learned was far beyond the Analytics Cloud content, which is why I primarily took the class.
The course hands-on exercises with free trial accounts along with OpenSAP (Thanks, SAP!) is very powerful to learn and demonstrate the fundamental capabilities to others. I only wish the course had been longer with even more in-depth exercises, but it provided all the fundamentals for those of us curious to experiment. I plan to get another SCP 30-day trial later with a different email address to continue evaluating capabilities. 30 days just isn't enough for "spare-time" learning!
On SDI, I've used a different 3-part blog on deploying SDI in SCP along with the DP agent using the flat file adapter. I put the DP adapter on my own, remote Win10 virtual machine where I can easily edit the flat files, model & query the data via SCP, and now with Analytics Cloud or the other services in SCP. Using SCP and SDI makes for an easy yet compelling demo of why SCP is the center of an "intelligent enterprise", particularly for analytics. When the SAC live models get the Smart Discovery capabilities (currently requires acquired models), it'll be awesome!
These classes and trial systems (thanks again SAP!) have really helped me understand literally what's possible today. This course was especially useful for Analytics!
Thanks Phil for the great blog -- and what a great comment, Doug! You might want to consider a blog post yourself to share this information more widely 🙂
Thanks Moya Watson - appreciate the feedback and yeah totally agree - Doug's reply was awesome!
Great reply mate and thankyou for sharing your experiences on the course - this is exactly why I wrote the blog post. It is very interesting to hear what others get out of the course and the experiences they've had with the Analytics capabilities here. Agree with your feedback around more detailed exercises - I know there are blogs out there to install SDI etc but would have been good to do this as part of the course.
Thanks for sharing the SDI tutorials, I'm thankful there are other community members that can add to the conversation - much appreciated! I will check them out to deepen my understanding of SDI and how it can be used.